Project
Loading the data
carbonData<-read.csv('/Users/angadsingh/Downloads/Carbon Emission.csv')
summary(carbonData) Body.Type Sex Diet How.Often.Shower Heating.Energy.Source Transport
Length:10000 Length:10000 Length:10000 Length:10000 Length:10000 Length:10000
Class :character Class :character Class :character Class :character Class :character Class :character
Mode :character Mode :character Mode :character Mode :character Mode :character Mode :character
Vehicle.Type Social.Activity Monthly.Grocery.Bill Frequency.of.Traveling.by.Air Vehicle.Monthly.Distance.Km
Length:10000 Length:10000 Min. : 50.0 Length:10000 Min. : 0
Class :character Class :character 1st Qu.:111.0 Class :character 1st Qu.: 69
Mode :character Mode :character Median :173.0 Mode :character Median : 823
Mean :173.9 Mean :2031
3rd Qu.:237.0 3rd Qu.:2517
Max. :299.0 Max. :9999
Waste.Bag.Size Waste.Bag.Weekly.Count How.Long.TV.PC.Daily.Hour How.Many.New.Clothes.Monthly How.Long.Internet.Daily.Hour
Length:10000 Min. :1.000 Min. : 0.00 Min. : 0.00 Min. : 0.00
Class :character 1st Qu.:2.000 1st Qu.: 6.00 1st Qu.:13.00 1st Qu.: 6.00
Mode :character Median :4.000 Median :12.00 Median :25.00 Median :12.00
Mean :4.025 Mean :12.14 Mean :25.11 Mean :11.89
3rd Qu.:6.000 3rd Qu.:18.00 3rd Qu.:38.00 3rd Qu.:18.00
Max. :7.000 Max. :24.00 Max. :50.00 Max. :24.00
Energy.efficiency Recycling Cooking_With CarbonEmission
Length:10000 Length:10000 Length:10000 Min. : 306
Class :character Class :character Class :character 1st Qu.:1538
Mode :character Mode :character Mode :character Median :2080
Mean :2269
3rd Qu.:2768
Max. :8377 str(carbonData)'data.frame': 10000 obs. of 20 variables:
$ Body.Type : chr "overweight" "obese" "overweight" "overweight" ...
$ Sex : chr "female" "female" "male" "male" ...
$ Diet : chr "pescatarian" "vegetarian" "omnivore" "omnivore" ...
$ How.Often.Shower : chr "daily" "less frequently" "more frequently" "twice a day" ...
$ Heating.Energy.Source : chr "coal" "natural gas" "wood" "wood" ...
$ Transport : chr "public" "walk/bicycle" "private" "walk/bicycle" ...
$ Vehicle.Type : chr "" "" "petrol" "" ...
$ Social.Activity : chr "often" "often" "never" "sometimes" ...
$ Monthly.Grocery.Bill : int 230 114 138 157 266 144 56 59 200 135 ...
$ Frequency.of.Traveling.by.Air: chr "frequently" "rarely" "never" "rarely" ...
$ Vehicle.Monthly.Distance.Km : int 210 9 2472 74 8457 658 5363 54 1376 440 ...
$ Waste.Bag.Size : chr "large" "extra large" "small" "medium" ...
$ Waste.Bag.Weekly.Count : int 4 3 1 3 1 1 4 3 3 1 ...
$ How.Long.TV.PC.Daily.Hour : int 7 9 14 20 3 22 9 5 3 8 ...
$ How.Many.New.Clothes.Monthly : int 26 38 47 5 5 18 11 39 31 23 ...
$ How.Long.Internet.Daily.Hour : int 1 5 6 7 6 9 19 15 15 18 ...
$ Energy.efficiency : chr "No" "No" "Sometimes" "Sometimes" ...
$ Recycling : chr "['Metal']" "['Metal']" "['Metal']" "['Paper', 'Plastic', 'Glass', 'Metal']" ...
$ Cooking_With : chr "['Stove', 'Oven']" "['Stove', 'Microwave']" "['Oven', 'Microwave']" "['Microwave', 'Grill', 'Airfryer']" ...
$ CarbonEmission : int 2238 1892 2595 1074 4743 1647 1832 2322 2494 1178 ...From the str of carbon data i can see that i am having empty vehicle types as “” so i will replace them with No vehicle
carbonData$Vehicle.Type[carbonData$Transport=='public'|carbonData$Transport=='walk/bicycle']<-'FuelEfficient'
#carbonData<- carbonData %>% mutate(Vehicle.Type=ifelse(Vehicle.Type=="","No vehicle",Vehicle.Type))
str(carbonData)'data.frame': 10000 obs. of 20 variables:
$ Body.Type : chr "overweight" "obese" "overweight" "overweight" ...
$ Sex : chr "female" "female" "male" "male" ...
$ Diet : chr "pescatarian" "vegetarian" "omnivore" "omnivore" ...
$ How.Often.Shower : chr "daily" "less frequently" "more frequently" "twice a day" ...
$ Heating.Energy.Source : chr "coal" "natural gas" "wood" "wood" ...
$ Transport : chr "public" "walk/bicycle" "private" "walk/bicycle" ...
$ Vehicle.Type : chr "FuelEfficient" "FuelEfficient" "petrol" "FuelEfficient" ...
$ Social.Activity : chr "often" "often" "never" "sometimes" ...
$ Monthly.Grocery.Bill : int 230 114 138 157 266 144 56 59 200 135 ...
$ Frequency.of.Traveling.by.Air: chr "frequently" "rarely" "never" "rarely" ...
$ Vehicle.Monthly.Distance.Km : int 210 9 2472 74 8457 658 5363 54 1376 440 ...
$ Waste.Bag.Size : chr "large" "extra large" "small" "medium" ...
$ Waste.Bag.Weekly.Count : int 4 3 1 3 1 1 4 3 3 1 ...
$ How.Long.TV.PC.Daily.Hour : int 7 9 14 20 3 22 9 5 3 8 ...
$ How.Many.New.Clothes.Monthly : int 26 38 47 5 5 18 11 39 31 23 ...
$ How.Long.Internet.Daily.Hour : int 1 5 6 7 6 9 19 15 15 18 ...
$ Energy.efficiency : chr "No" "No" "Sometimes" "Sometimes" ...
$ Recycling : chr "['Metal']" "['Metal']" "['Metal']" "['Paper', 'Plastic', 'Glass', 'Metal']" ...
$ Cooking_With : chr "['Stove', 'Oven']" "['Stove', 'Microwave']" "['Oven', 'Microwave']" "['Microwave', 'Grill', 'Airfryer']" ...
$ CarbonEmission : int 2238 1892 2595 1074 4743 1647 1832 2322 2494 1178 ...#carbonData[carbonData == ""]<-NA
colSums(is.na(carbonData)) Body.Type Sex Diet How.Often.Shower
0 0 0 0
Heating.Energy.Source Transport Vehicle.Type Social.Activity
0 0 0 0
Monthly.Grocery.Bill Frequency.of.Traveling.by.Air Vehicle.Monthly.Distance.Km Waste.Bag.Size
0 0 0 0
Waste.Bag.Weekly.Count How.Long.TV.PC.Daily.Hour How.Many.New.Clothes.Monthly How.Long.Internet.Daily.Hour
0 0 0 0
Energy.efficiency Recycling Cooking_With CarbonEmission
0 0 0 0 library(dplyr)
carbonData<-carbonData %>%
mutate_if(is.character, as.factor)%>%
mutate_if(is.integer, as.numeric)
str(carbonData)'data.frame': 10000 obs. of 20 variables:
$ Body.Type : Factor w/ 4 levels "normal","obese",..: 3 2 3 3 2 3 4 4 3 4 ...
$ Sex : Factor w/ 2 levels "female","male": 1 1 2 2 1 2 1 1 2 1 ...
$ Diet : Factor w/ 4 levels "omnivore","pescatarian",..: 2 4 1 1 4 4 3 3 1 2 ...
$ How.Often.Shower : Factor w/ 4 levels "daily","less frequently",..: 1 2 3 4 1 2 2 3 1 1 ...
$ Heating.Energy.Source : Factor w/ 4 levels "coal","electricity",..: 1 3 4 4 1 4 4 1 4 4 ...
$ Transport : Factor w/ 3 levels "private","public",..: 2 3 1 3 1 2 1 3 2 2 ...
$ Vehicle.Type : Factor w/ 6 levels "diesel","electric",..: 3 3 6 3 1 3 4 3 3 3 ...
$ Social.Activity : Factor w/ 3 levels "never","often",..: 2 2 1 3 2 3 1 3 1 2 ...
$ Monthly.Grocery.Bill : num 230 114 138 157 266 144 56 59 200 135 ...
$ Frequency.of.Traveling.by.Air: Factor w/ 4 levels "frequently","never",..: 1 3 2 3 4 1 3 4 1 3 ...
$ Vehicle.Monthly.Distance.Km : num 210 9 2472 74 8457 ...
$ Waste.Bag.Size : Factor w/ 4 levels "extra large",..: 2 1 4 3 2 2 3 1 3 1 ...
$ Waste.Bag.Weekly.Count : num 4 3 1 3 1 1 4 3 3 1 ...
$ How.Long.TV.PC.Daily.Hour : num 7 9 14 20 3 22 9 5 3 8 ...
$ How.Many.New.Clothes.Monthly : num 26 38 47 5 5 18 11 39 31 23 ...
$ How.Long.Internet.Daily.Hour : num 1 5 6 7 6 9 19 15 15 18 ...
$ Energy.efficiency : Factor w/ 3 levels "No","Sometimes",..: 1 1 2 2 3 2 2 1 3 2 ...
$ Recycling : Factor w/ 16 levels "['Glass', 'Metal']",..: 3 3 3 7 11 4 16 8 2 2 ...
$ Cooking_With : Factor w/ 16 levels "['Grill', 'Airfryer']",..: 14 10 6 2 7 13 1 10 2 2 ...
$ CarbonEmission : num 2238 1892 2595 1074 4743 ...summary(carbonData) Body.Type Sex Diet How.Often.Shower Heating.Energy.Source Transport
normal :2473 female:5007 omnivore :2492 daily :2546 coal :2523 private :3279
obese :2500 male :4993 pescatarian:2554 less frequently:2487 electricity:2552 public :3294
overweight :2487 vegan :2497 more frequently:2451 natural gas:2462 walk/bicycle:3427
underweight:2540 vegetarian :2457 twice a day :2516 wood :2463
Vehicle.Type Social.Activity Monthly.Grocery.Bill Frequency.of.Traveling.by.Air Vehicle.Monthly.Distance.Km
diesel : 622 never :3406 Min. : 50.0 frequently :2524 Min. : 0
electric : 671 often :3319 1st Qu.:111.0 never :2459 1st Qu.: 69
FuelEfficient:6721 sometimes:3275 Median :173.0 rarely :2477 Median : 823
hybrid : 642 Mean :173.9 very frequently:2540 Mean :2031
lpg : 697 3rd Qu.:237.0 3rd Qu.:2517
petrol : 647 Max. :299.0 Max. :9999
Waste.Bag.Size Waste.Bag.Weekly.Count How.Long.TV.PC.Daily.Hour How.Many.New.Clothes.Monthly How.Long.Internet.Daily.Hour
extra large:2500 Min. :1.000 Min. : 0.00 Min. : 0.00 Min. : 0.00
large :2501 1st Qu.:2.000 1st Qu.: 6.00 1st Qu.:13.00 1st Qu.: 6.00
medium :2474 Median :4.000 Median :12.00 Median :25.00 Median :12.00
small :2525 Mean :4.025 Mean :12.14 Mean :25.11 Mean :11.89
3rd Qu.:6.000 3rd Qu.:18.00 3rd Qu.:38.00 3rd Qu.:18.00
Max. :7.000 Max. :24.00 Max. :50.00 Max. :24.00
Energy.efficiency Recycling Cooking_With
No :3221 [] : 675 ['Stove', 'Oven'] : 670
Sometimes:3463 ['Paper', 'Plastic', 'Metal'] : 648 ['Stove', 'Microwave', 'Grill', 'Airfryer'] : 652
Yes :3316 ['Paper', 'Glass', 'Metal'] : 647 ['Oven', 'Microwave'] : 649
['Glass', 'Metal'] : 645 ['Oven', 'Microwave', 'Grill', 'Airfryer'] : 638
['Paper', 'Plastic', 'Glass', 'Metal']: 637 ['Stove', 'Oven', 'Microwave', 'Grill', 'Airfryer']: 637
['Paper', 'Plastic'] : 633 ['Stove', 'Grill', 'Airfryer'] : 628
(Other) :6115 (Other) :6126
CarbonEmission
Min. : 306
1st Qu.:1538
Median :2080
Mean :2269
3rd Qu.:2768
Max. :8377
table(carbonData$Body.Type)
normal obese overweight underweight
2473 2500 2487 2540 table(carbonData$Sex)
female male
5007 4993 table(carbonData$Diet)
omnivore pescatarian vegan vegetarian
2492 2554 2497 2457 table(carbonData$How.Often.Shower)
daily less frequently more frequently twice a day
2546 2487 2451 2516 table(carbonData$Heating.Energy.Source)
coal electricity natural gas wood
2523 2552 2462 2463 table(carbonData$Transport)
private public walk/bicycle
3279 3294 3427 table(carbonData$Social.Activity)
never often sometimes
3406 3319 3275 table(carbonData$Frequency.of.Traveling.by.Air)
frequently never rarely very frequently
2524 2459 2477 2540 table(carbonData$Waste.Bag.Size)
extra large large medium small
2500 2501 2474 2525 table(carbonData$Energy.efficiency)
No Sometimes Yes
3221 3463 3316 hist(carbonData$CarbonEmission)
carbonData$CarbonEmission<-log(carbonData$CarbonEmission)
hist(carbonData$CarbonEmission)
carbonIndices<-which(names(carbonData)=='CarbonEmission')
for (c in colnames(carbonData[,-carbonIndices])) {
if(is.factor(carbonData[,c])){
try({
anovaaResult<-aov(carbonData$CarbonEmission~carbonData[,c])
cat("ANOVA of ",c, "and CarbonEmission", "\n")
print(summary(anovaaResult))
boxplot(carbonData$CarbonEmission~carbonData[,c],shade=TRUE, main = paste("Carbon Emission vs", c), xlab ="CarbonEmission", ylab=c ,col="lightgreen")
})
}
else if (is.numeric(carbonData[,c])){
try({
corTest<-cor.test(carbonData$CarbonEmission,carbonData[,c], method = "pearson")
cat("p.value of ",c, "and Carbon Emission", corTest$p.value, "\n")
plot(carbonData$CarbonEmission,carbonData[,c], main = paste("Carbon Emission vs", c), xlab ="Carbon Emission", ylab=c)
})
}
}ANOVA of Body.Type and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 82.7 27.583 149 <2e-16 ***
Residuals 9996 1850.3 0.185
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Sex and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 1 55.5 55.51 295.6 <2e-16 ***
Residuals 9998 1877.5 0.19
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Diet and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 11.8 3.944 20.52 2.98e-13 ***
Residuals 9996 1921.2 0.192
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of How.Often.Shower and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 1.9 0.6333 3.278 0.0201 *
Residuals 9996 1931.1 0.1932
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Heating.Energy.Source and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 71.1 23.702 127.2 <2e-16 ***
Residuals 9996 1861.9 0.186
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Transport and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 2 384.7 192.34 1242 <2e-16 ***
Residuals 9997 1548.3 0.15
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Vehicle.Type and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 5 573.6 114.71 843.3 <2e-16 ***
Residuals 9994 1359.5 0.14
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Social.Activity and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 2 8.5 4.248 22.07 2.74e-10 ***
Residuals 9997 1924.5 0.193
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
p.value of Monthly.Grocery.Bill and Carbon Emission 8.380793e-21 
ANOVA of Frequency.of.Traveling.by.Air and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 568.2 189.41 1387 <2e-16 ***
Residuals 9996 1364.8 0.14
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
p.value of Vehicle.Monthly.Distance.Km and Carbon Emission 0 
ANOVA of Waste.Bag.Size and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 3 53.1 17.692 94.07 <2e-16 ***
Residuals 9996 1879.9 0.188
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
p.value of Waste.Bag.Weekly.Count and Carbon Emission 3.257926e-79 
p.value of How.Long.TV.PC.Daily.Hour and Carbon Emission 0.2282313 
p.value of How.Many.New.Clothes.Monthly and Carbon Emission 8.851718e-131 
p.value of How.Long.Internet.Daily.Hour and Carbon Emission 1.411988e-09 
ANOVA of Energy.efficiency and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 2 0.7 0.3284 1.699 0.183
Residuals 9997 1932.4 0.1933 
ANOVA of Recycling and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 15 41.7 2.7802 14.68 <2e-16 ***
Residuals 9984 1891.3 0.1894
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
ANOVA of Cooking_With and CarbonEmission
Df Sum Sq Mean Sq F value Pr(>F)
carbonData[, c] 15 5 0.3301 1.709 0.0422 *
Residuals 9984 1928 0.1931
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
library(tidyverse)── Attaching core tidyverse packages ───────────────────────────────────────────────────────────────────────────── tidyverse 2.0.0 ──
✔ forcats 1.0.0 ✔ stringr 1.5.1
✔ lubridate 1.9.3 ✔ tibble 3.2.1
✔ purrr 1.0.2 ✔ tidyr 1.3.1
✔ readr 2.1.5 ── Conflicts ─────────────────────────────────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag() masks stats::lag()
✖ purrr::lift() masks caret::lift()
ℹ Use the ]8;;http://conflicted.r-lib.org/conflicted package]8;; to force all conflicts to become errorsparseList<-function(x){
str_remove_all(x,"\\[|\\]|'")%>%
strsplit(", ")%>%
unlist()
}
carbonData$Recycling<-sapply(carbonData$Recycling,parseList)
carbonData$Cooking_With<-sapply(carbonData$Cooking_With,parseList)
carbonData$Recycling<-sapply(carbonData$Recycling,paste,collapse=",")
carbonData$Cooking_With<-sapply(carbonData$Cooking_With,paste,collapse=",")
#str(carbonData)
dummies<-function(col){
items<-unlist(str_split(col,","))
items<-trimws(items)
items<-items[items != ""]
uniqueItems<-unique(items)
dummyDataFrame<-data.frame(matrix(0,nrow = length(col),ncol = length(uniqueItems)))
colnames(dummyDataFrame)<-uniqueItems
for (i in seq_along(col)) {
rowItems<-unlist(str_split(col[i],","))%>%
map_chr(~str_trim(.))%>%
discard(~.=="")
rowItems<-rowItems[rowItems %in% uniqueItems]
dummyDataFrame[i,rowItems]<-1
}
return(dummyDataFrame)
}
recyclingDummies<-dummies(carbonData$Recycling)
cookingDummies<-dummies(carbonData$Cooking_With)
carbonData<-cbind(carbonData,recyclingDummies,cookingDummies)
carbonData$Recycling<- NULL
carbonData$Cooking_With<-NULL
str(carbonData)'data.frame': 10000 obs. of 27 variables:
$ Body.Type : Factor w/ 4 levels "normal","obese",..: 3 2 3 3 2 3 4 4 3 4 ...
$ Sex : Factor w/ 2 levels "female","male": 1 1 2 2 1 2 1 1 2 1 ...
$ Diet : Factor w/ 4 levels "omnivore","pescatarian",..: 2 4 1 1 4 4 3 3 1 2 ...
$ How.Often.Shower : Factor w/ 4 levels "daily","less frequently",..: 1 2 3 4 1 2 2 3 1 1 ...
$ Heating.Energy.Source : Factor w/ 4 levels "coal","electricity",..: 1 3 4 4 1 4 4 1 4 4 ...
$ Transport : Factor w/ 3 levels "private","public",..: 2 3 1 3 1 2 1 3 2 2 ...
$ Vehicle.Type : Factor w/ 6 levels "diesel","electric",..: 3 3 6 3 1 3 4 3 3 3 ...
$ Social.Activity : Factor w/ 3 levels "never","often",..: 2 2 1 3 2 3 1 3 1 2 ...
$ Monthly.Grocery.Bill : num 230 114 138 157 266 144 56 59 200 135 ...
$ Frequency.of.Traveling.by.Air: Factor w/ 4 levels "frequently","never",..: 1 3 2 3 4 1 3 4 1 3 ...
$ Vehicle.Monthly.Distance.Km : num 210 9 2472 74 8457 ...
$ Waste.Bag.Size : Factor w/ 4 levels "extra large",..: 2 1 4 3 2 2 3 1 3 1 ...
$ Waste.Bag.Weekly.Count : num 4 3 1 3 1 1 4 3 3 1 ...
$ How.Long.TV.PC.Daily.Hour : num 7 9 14 20 3 22 9 5 3 8 ...
$ How.Many.New.Clothes.Monthly : num 26 38 47 5 5 18 11 39 31 23 ...
$ How.Long.Internet.Daily.Hour : num 1 5 6 7 6 9 19 15 15 18 ...
$ Energy.efficiency : Factor w/ 3 levels "No","Sometimes",..: 1 1 2 2 3 2 2 1 3 2 ...
$ CarbonEmission : num 7.71 7.55 7.86 6.98 8.46 ...
$ Metal : num 1 1 1 1 0 1 0 0 0 0 ...
$ Paper : num 0 0 0 1 1 1 0 1 0 0 ...
$ Plastic : num 0 0 0 1 0 0 0 1 0 0 ...
$ Glass : num 0 0 0 1 0 1 0 1 1 1 ...
$ Stove : num 1 1 0 0 0 1 0 1 0 0 ...
$ Oven : num 1 0 1 0 1 1 0 0 0 0 ...
$ Microwave : num 0 1 1 1 0 1 0 1 1 1 ...
$ Grill : num 0 0 0 1 0 0 1 0 1 1 ...
$ Airfryer : num 0 0 0 1 0 0 1 0 1 1 ...library(caret)
carbonDataIndexs <- createDataPartition(carbonData$CarbonEmission, p=0.8, list=FALSE)
carbonTrainData<-carbonData[carbonDataIndexs,]
carbonTrainData
carbonTestData<-carbonData[-carbonDataIndexs,]
carbonTestData
carbonTestLabels<-carbonTestData$CarbonEmission#Benchmark
meanTransport<- carbonTrainData %>% #this will calculate emission per transport type
group_by(Transport) %>%
summarize(meanEmission= mean(CarbonEmission, na.rm = TRUE))
meanTransportTable<-setNames(meanTransport$meanEmission,meanTransport$Transport) # will store transport levels to its average emission
predictTransport<-function(row){
transportType<-as.character(row["Transport"])
if(transportType %in% names(meanTransportTable)){
return(meanTransportTable[[transportType]])
}
}
benchmarkPred<-apply(carbonTestData,1,predictTransport)
rmse(benchmarkPred,carbonTestData$CarbonEmission)[1] 0.004286257MAE(benchmarkPred,carbonTestData$CarbonEmission)[1] 0.3169533rsquaredNew<-sum((benchmarkPred-carbonTestData$CarbonEmission)^2)/sum((carbonTestData$CarbonEmission-mean(carbonTestData$CarbonEmission))^2)
rsquaredNew[1] 0.8106951knnModel<-train(CarbonEmission~.,data = carbonTrainData, method="knn", trControl=trainControl(method = "cv", number=5))knnModelk-Nearest Neighbors
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6400, 6401, 6401
Resampling results across tuning parameters:
k RMSE Rsquared MAE
5 0.3912082 0.2416923 0.3099291
7 0.3820210 0.2612955 0.3027079
9 0.3789765 0.2674213 0.3010474
RMSE was used to select the optimal model using the smallest value.
The final value used for the model was k = 9.knnPred<-predict(knnModel,newdata = carbonTestData)
rmse=function(x,y){
return((mean(x-y)^2))
}
rmse(knnPred,carbonTestLabels)[1] 1.626558e-07lmModel<-train(CarbonEmission~.,data = carbonTrainData, method="lm", trControl=trainControl(method = "cv", number=5))
lmModelLinear Regression
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results:
RMSE Rsquared MAE
0.1216956 0.9235284 0.08674322
Tuning parameter 'intercept' was held constant at a value of TRUEsummary(lmModel)
Call:
lm(formula = .outcome ~ ., data = dat)
Residuals:
Min 1Q Median 3Q Max
-0.77905 -0.05477 0.00826 0.06510 0.47201
Coefficients: (2 not defined because of singularities)
Estimate Std. Error t value Pr(>|t|)
(Intercept) 7.409e+00 1.167e-02 634.830 < 2e-16 ***
Body.Typeobese 1.843e-01 3.854e-03 47.825 < 2e-16 ***
Body.Typeoverweight 9.380e-02 3.854e-03 24.337 < 2e-16 ***
Body.Typeunderweight -4.979e-02 3.841e-03 -12.964 < 2e-16 ***
Sexmale 1.511e-01 2.721e-03 55.551 < 2e-16 ***
Dietpescatarian -4.430e-02 3.843e-03 -11.529 < 2e-16 ***
Dietvegan -7.990e-02 3.860e-03 -20.701 < 2e-16 ***
Dietvegetarian -7.180e-02 3.871e-03 -18.548 < 2e-16 ***
`How.Often.Showerless frequently` -7.940e-03 3.827e-03 -2.075 0.03803 *
`How.Often.Showermore frequently` 1.786e-02 3.850e-03 4.638 3.57e-06 ***
`How.Often.Showertwice a day` 1.058e-02 3.838e-03 2.755 0.00588 **
Heating.Energy.Sourceelectricity -2.236e-01 3.809e-03 -58.694 < 2e-16 ***
`Heating.Energy.Sourcenatural gas` -9.702e-02 3.843e-03 -25.244 < 2e-16 ***
Heating.Energy.Sourcewood -9.863e-02 3.866e-03 -25.513 < 2e-16 ***
Transportpublic -1.970e-01 6.834e-03 -28.834 < 2e-16 ***
`Transportwalk/bicycle` -1.722e-01 7.231e-03 -23.809 < 2e-16 ***
Vehicle.Typeelectric -5.033e-01 7.590e-03 -66.310 < 2e-16 ***
Vehicle.TypeFuelEfficient NA NA NA NA
Vehicle.Typehybrid -1.377e-01 7.638e-03 -18.032 < 2e-16 ***
Vehicle.Typelpg 3.441e-02 7.609e-03 4.523 6.19e-06 ***
Vehicle.Typepetrol 1.909e-01 7.639e-03 24.994 < 2e-16 ***
Social.Activityoften 8.625e-02 3.329e-03 25.912 < 2e-16 ***
Social.Activitysometimes 3.894e-02 3.326e-03 11.707 < 2e-16 ***
Monthly.Grocery.Bill 4.704e-04 1.873e-05 25.120 < 2e-16 ***
Frequency.of.Traveling.by.Airnever -3.610e-01 3.865e-03 -93.387 < 2e-16 ***
Frequency.of.Traveling.by.Airrarely -2.413e-01 3.838e-03 -62.885 < 2e-16 ***
`Frequency.of.Traveling.by.Airvery frequently` 2.643e-01 3.815e-03 69.276 < 2e-16 ***
Vehicle.Monthly.Distance.Km 6.730e-05 8.023e-07 83.884 < 2e-16 ***
Waste.Bag.Sizelarge -6.017e-02 3.833e-03 -15.698 < 2e-16 ***
Waste.Bag.Sizemedium -1.264e-01 3.841e-03 -32.917 < 2e-16 ***
Waste.Bag.Sizesmall -1.948e-01 3.849e-03 -50.609 < 2e-16 ***
Waste.Bag.Weekly.Count 4.161e-02 6.827e-04 60.946 < 2e-16 ***
How.Long.TV.PC.Daily.Hour 1.272e-03 1.909e-04 6.662 2.89e-11 ***
How.Many.New.Clothes.Monthly 7.067e-03 9.244e-05 76.452 < 2e-16 ***
How.Long.Internet.Daily.Hour 3.958e-03 1.869e-04 21.177 < 2e-16 ***
Energy.efficiencySometimes -2.072e-02 3.322e-03 -6.236 4.73e-10 ***
Energy.efficiencyYes -3.183e-02 3.370e-03 -9.444 < 2e-16 ***
Metal -6.951e-02 2.722e-03 -25.540 < 2e-16 ***
Paper -7.439e-02 2.720e-03 -27.347 < 2e-16 ***
Plastic -2.879e-02 2.722e-03 -10.575 < 2e-16 ***
Glass -4.867e-02 2.718e-03 -17.907 < 2e-16 ***
Stove 1.493e-02 2.719e-03 5.492 4.10e-08 ***
Oven 1.834e-02 2.721e-03 6.743 1.67e-11 ***
Microwave 7.785e-03 2.719e-03 2.863 0.00420 **
Grill 1.791e-02 2.719e-03 6.589 4.72e-11 ***
Airfryer NA NA NA NA
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 0.1213 on 7957 degrees of freedom
Multiple R-squared: 0.9243, Adjusted R-squared: 0.9239
F-statistic: 2260 on 43 and 7957 DF, p-value: < 2.2e-16stepwiseModel<-train(CarbonEmission~.,data = carbonTrainData, method="leapBackward", trControl=trainControl(method = "cv", number=5))Warning: 2 linear dependencies foundReordering variables and trying again:Warning: 2 linear dependencies foundReordering variables and trying again:Warning: 2 linear dependencies foundReordering variables and trying again:Warning: 2 linear dependencies foundReordering variables and trying again:Warning: 2 linear dependencies foundReordering variables and trying again:Warning: 2 linear dependencies foundReordering variables and trying again:stepwiseModelLinear Regression with Backwards Selection
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6400, 6401, 6401
Resampling results across tuning parameters:
nvmax RMSE Rsquared MAE
2 0.3760164 0.2691345 0.2992884
3 0.3594632 0.3318216 0.2867232
4 0.3460790 0.3807424 0.2727536
RMSE was used to select the optimal model using the smallest value.
The final value used for the model was nvmax = 4.summary(stepwiseModel$finalModel)Subset selection object
45 Variables (and intercept)
Forced in Forced out
Body.Typeobese FALSE FALSE
Body.Typeoverweight FALSE FALSE
Body.Typeunderweight FALSE FALSE
Sexmale FALSE FALSE
Dietpescatarian FALSE FALSE
Dietvegan FALSE FALSE
Dietvegetarian FALSE FALSE
How.Often.Showerless frequently FALSE FALSE
How.Often.Showermore frequently FALSE FALSE
How.Often.Showertwice a day FALSE FALSE
Heating.Energy.Sourceelectricity FALSE FALSE
Heating.Energy.Sourcenatural gas FALSE FALSE
Heating.Energy.Sourcewood FALSE FALSE
Transportpublic FALSE FALSE
Transportwalk/bicycle FALSE FALSE
Vehicle.Typeelectric FALSE FALSE
Vehicle.Typehybrid FALSE FALSE
Vehicle.Typelpg FALSE FALSE
Vehicle.Typepetrol FALSE FALSE
Social.Activityoften FALSE FALSE
Social.Activitysometimes FALSE FALSE
Monthly.Grocery.Bill FALSE FALSE
Frequency.of.Traveling.by.Airnever FALSE FALSE
Frequency.of.Traveling.by.Airrarely FALSE FALSE
Frequency.of.Traveling.by.Airvery frequently FALSE FALSE
Vehicle.Monthly.Distance.Km FALSE FALSE
Waste.Bag.Sizelarge FALSE FALSE
Waste.Bag.Sizemedium FALSE FALSE
Waste.Bag.Sizesmall FALSE FALSE
Waste.Bag.Weekly.Count FALSE FALSE
How.Long.TV.PC.Daily.Hour FALSE FALSE
How.Many.New.Clothes.Monthly FALSE FALSE
How.Long.Internet.Daily.Hour FALSE FALSE
Energy.efficiencySometimes FALSE FALSE
Energy.efficiencyYes FALSE FALSE
Metal FALSE FALSE
Paper FALSE FALSE
Plastic FALSE FALSE
Glass FALSE FALSE
Stove FALSE FALSE
Oven FALSE FALSE
Microwave FALSE FALSE
Grill FALSE FALSE
Vehicle.TypeFuelEfficient FALSE FALSE
Airfryer FALSE FALSE
1 subsets of each size up to 5
Selection Algorithm: backward
Body.Typeobese Body.Typeoverweight Body.Typeunderweight Sexmale Dietpescatarian Dietvegan Dietvegetarian How.Often.Showerless frequently
1 ( 1 ) " " " " " " " " " " " " " " " "
2 ( 1 ) " " " " " " " " " " " " " " " "
3 ( 1 ) " " " " " " " " " " " " " " " "
4 ( 1 ) " " " " " " " " " " " " " " " "
5 ( 1 ) " " " " " " " " " " " " " " " "
How.Often.Showermore frequently How.Often.Showertwice a day Heating.Energy.Sourceelectricity Heating.Energy.Sourcenatural gas
1 ( 1 ) " " " " " " " "
2 ( 1 ) " " " " " " " "
3 ( 1 ) " " " " " " " "
4 ( 1 ) " " " " " " " "
5 ( 1 ) " " " " " " " "
Heating.Energy.Sourcewood Transportpublic Transportwalk/bicycle Vehicle.Typeelectric Vehicle.TypeFuelEfficient Vehicle.Typehybrid Vehicle.Typelpg
1 ( 1 ) " " " " " " " " " " " " " "
2 ( 1 ) " " " " " " " " " " " " " "
3 ( 1 ) " " " " " " "*" " " " " " "
4 ( 1 ) " " " " " " "*" " " " " " "
5 ( 1 ) " " " " " " "*" " " " " " "
Vehicle.Typepetrol Social.Activityoften Social.Activitysometimes Monthly.Grocery.Bill Frequency.of.Traveling.by.Airnever
1 ( 1 ) " " " " " " " " " "
2 ( 1 ) " " " " " " " " " "
3 ( 1 ) " " " " " " " " " "
4 ( 1 ) " " " " " " " " " "
5 ( 1 ) " " " " " " " " "*"
Frequency.of.Traveling.by.Airrarely Frequency.of.Traveling.by.Airvery frequently Vehicle.Monthly.Distance.Km Waste.Bag.Sizelarge
1 ( 1 ) " " " " "*" " "
2 ( 1 ) " " "*" "*" " "
3 ( 1 ) " " "*" "*" " "
4 ( 1 ) " " "*" "*" " "
5 ( 1 ) " " "*" "*" " "
Waste.Bag.Sizemedium Waste.Bag.Sizesmall Waste.Bag.Weekly.Count How.Long.TV.PC.Daily.Hour How.Many.New.Clothes.Monthly How.Long.Internet.Daily.Hour
1 ( 1 ) " " " " " " " " " " " "
2 ( 1 ) " " " " " " " " " " " "
3 ( 1 ) " " " " " " " " " " " "
4 ( 1 ) " " " " " " " " "*" " "
5 ( 1 ) " " " " " " " " "*" " "
Energy.efficiencySometimes Energy.efficiencyYes Metal Paper Plastic Glass Stove Oven Microwave Grill Airfryer
1 ( 1 ) " " " " " " " " " " " " " " " " " " " " " "
2 ( 1 ) " " " " " " " " " " " " " " " " " " " " " "
3 ( 1 ) " " " " " " " " " " " " " " " " " " " " " "
4 ( 1 ) " " " " " " " " " " " " " " " " " " " " " "
5 ( 1 ) " " " " " " " " " " " " " " " " " " " " " " colSums(is.na(carbonTrainData)) Body.Type Sex Diet How.Often.Shower Heating.Energy.Source
0 0 0 0 0
Transport Vehicle.Type Social.Activity Monthly.Grocery.Bill Frequency.of.Traveling.by.Air
0 0 0 0 0
Vehicle.Monthly.Distance.Km Waste.Bag.Size Waste.Bag.Weekly.Count How.Long.TV.PC.Daily.Hour How.Many.New.Clothes.Monthly
0 0 0 0 0
How.Long.Internet.Daily.Hour Energy.efficiency CarbonEmission Metal Paper
0 0 0 0 0
Plastic Glass Stove Oven Microwave
0 0 0 0 0
Grill Airfryer
0 0 #Lasso Model
library(glmnet)Loading required package: Matrix
Attaching package: ‘Matrix’
The following objects are masked from ‘package:tidyr’:
expand, pack, unpack
Loaded glmnet 4.1-8set.seed(1)
lassoModel<-train(CarbonEmission~.,data = carbonTrainData,method="glmnet",trControl= trainControl(method = "cv", number=5), tuneGrid = expand.grid(alpha=1, lambda=10^seq(-3,3,length=100))) Warning: There were missing values in resampled performance measures.lassoModelglmnet
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
lambda RMSE Rsquared MAE
1.000000e-03 0.1219858 0.9232209 0.08693488
1.149757e-03 0.1220753 0.9231430 0.08700875
1.321941e-03 0.1221904 0.9230430 0.08710667
1.519911e-03 0.1223470 0.9229054 0.08723849
1.747528e-03 0.1225536 0.9227225 0.08741574
2.009233e-03 0.1228246 0.9224806 0.08765246
2.310130e-03 0.1231771 0.9221635 0.08796258
2.656088e-03 0.1236457 0.9217360 0.08837683
3.053856e-03 0.1242478 0.9211821 0.08890477
3.511192e-03 0.1250325 0.9204489 0.08960023
4.037017e-03 0.1260563 0.9194736 0.09051846
4.641589e-03 0.1273779 0.9181895 0.09171832
5.336699e-03 0.1290497 0.9165381 0.09323783
6.135907e-03 0.1311917 0.9143534 0.09516533
7.054802e-03 0.1337763 0.9116869 0.09744389
8.111308e-03 0.1369090 0.9083708 0.10017575
9.326033e-03 0.1403485 0.9047807 0.10315995
1.072267e-02 0.1442686 0.9006956 0.10650907
1.232847e-02 0.1491610 0.8953001 0.11072814
1.417474e-02 0.1546006 0.8892623 0.11535445
1.629751e-02 0.1598460 0.8840559 0.11952361
1.873817e-02 0.1659921 0.8778439 0.12436408
2.154435e-02 0.1736061 0.8694320 0.13042497
2.477076e-02 0.1830635 0.8576678 0.13811167
2.848036e-02 0.1936998 0.8436314 0.14679304
3.274549e-02 0.2049344 0.8287289 0.15594333
3.764936e-02 0.2172973 0.8115630 0.16606920
4.328761e-02 0.2310128 0.7912730 0.17738897
4.977024e-02 0.2477212 0.7609757 0.19110356
5.722368e-02 0.2674957 0.7164689 0.20734886
6.579332e-02 0.2889289 0.6589664 0.22517705
7.564633e-02 0.3086367 0.6018407 0.24138756
8.697490e-02 0.3233407 0.5674578 0.25346893
1.000000e-01 0.3380457 0.5345973 0.26560896
1.149757e-01 0.3519869 0.5106866 0.27729282
1.321941e-01 0.3667388 0.4921571 0.28996428
1.519911e-01 0.3835151 0.4705402 0.30453448
1.747528e-01 0.4025545 0.4493947 0.32115299
2.009233e-01 0.4247240 0.3409155 0.34036123
2.310130e-01 0.4397931 NaN 0.35168920
2.656088e-01 0.4397931 NaN 0.35168920
3.053856e-01 0.4397931 NaN 0.35168920
3.511192e-01 0.4397931 NaN 0.35168920
4.037017e-01 0.4397931 NaN 0.35168920
4.641589e-01 0.4397931 NaN 0.35168920
5.336699e-01 0.4397931 NaN 0.35168920
6.135907e-01 0.4397931 NaN 0.35168920
7.054802e-01 0.4397931 NaN 0.35168920
8.111308e-01 0.4397931 NaN 0.35168920
9.326033e-01 0.4397931 NaN 0.35168920
1.072267e+00 0.4397931 NaN 0.35168920
1.232847e+00 0.4397931 NaN 0.35168920
1.417474e+00 0.4397931 NaN 0.35168920
1.629751e+00 0.4397931 NaN 0.35168920
1.873817e+00 0.4397931 NaN 0.35168920
2.154435e+00 0.4397931 NaN 0.35168920
2.477076e+00 0.4397931 NaN 0.35168920
2.848036e+00 0.4397931 NaN 0.35168920
3.274549e+00 0.4397931 NaN 0.35168920
3.764936e+00 0.4397931 NaN 0.35168920
4.328761e+00 0.4397931 NaN 0.35168920
4.977024e+00 0.4397931 NaN 0.35168920
5.722368e+00 0.4397931 NaN 0.35168920
6.579332e+00 0.4397931 NaN 0.35168920
7.564633e+00 0.4397931 NaN 0.35168920
8.697490e+00 0.4397931 NaN 0.35168920
1.000000e+01 0.4397931 NaN 0.35168920
1.149757e+01 0.4397931 NaN 0.35168920
1.321941e+01 0.4397931 NaN 0.35168920
1.519911e+01 0.4397931 NaN 0.35168920
1.747528e+01 0.4397931 NaN 0.35168920
2.009233e+01 0.4397931 NaN 0.35168920
2.310130e+01 0.4397931 NaN 0.35168920
2.656088e+01 0.4397931 NaN 0.35168920
3.053856e+01 0.4397931 NaN 0.35168920
3.511192e+01 0.4397931 NaN 0.35168920
4.037017e+01 0.4397931 NaN 0.35168920
4.641589e+01 0.4397931 NaN 0.35168920
5.336699e+01 0.4397931 NaN 0.35168920
6.135907e+01 0.4397931 NaN 0.35168920
7.054802e+01 0.4397931 NaN 0.35168920
8.111308e+01 0.4397931 NaN 0.35168920
9.326033e+01 0.4397931 NaN 0.35168920
1.072267e+02 0.4397931 NaN 0.35168920
1.232847e+02 0.4397931 NaN 0.35168920
1.417474e+02 0.4397931 NaN 0.35168920
1.629751e+02 0.4397931 NaN 0.35168920
1.873817e+02 0.4397931 NaN 0.35168920
2.154435e+02 0.4397931 NaN 0.35168920
2.477076e+02 0.4397931 NaN 0.35168920
2.848036e+02 0.4397931 NaN 0.35168920
3.274549e+02 0.4397931 NaN 0.35168920
3.764936e+02 0.4397931 NaN 0.35168920
4.328761e+02 0.4397931 NaN 0.35168920
4.977024e+02 0.4397931 NaN 0.35168920
5.722368e+02 0.4397931 NaN 0.35168920
6.579332e+02 0.4397931 NaN 0.35168920
7.564633e+02 0.4397931 NaN 0.35168920
8.697490e+02 0.4397931 NaN 0.35168920
1.000000e+03 0.4397931 NaN 0.35168920
Tuning parameter 'alpha' was held constant at a value of 1
RMSE was used to select the optimal model using the smallest value.
The final values used for the model were alpha = 1 and lambda = 0.001.lassoLambda<-lassoModel$bestTune$lambda
lassoPredictor<- setdiff(names(carbonTrainData),"CarbonEmission")
lassoFinalModel<-glmnet(as.matrix(carbonTrainData[,lassoPredictor]),carbonTrainData[,"CarbonEmission"],alpha = 1,lambda = lassoLambda, family = "gaussian")Warning: NAs introduced by coercioncoeff<-coef(lassoFinalModel)
coeff27 x 1 sparse Matrix of class "dgCMatrix"
s0
(Intercept) 7.064707e+00
Body.Type .
Sex .
Diet .
How.Often.Shower .
Heating.Energy.Source .
Transport .
Vehicle.Type .
Social.Activity .
Monthly.Grocery.Bill 5.264763e-04
Frequency.of.Traveling.by.Air .
Vehicle.Monthly.Distance.Km 8.078680e-05
Waste.Bag.Size .
Waste.Bag.Weekly.Count 4.161160e-02
How.Long.TV.PC.Daily.Hour 1.146001e-03
How.Many.New.Clothes.Monthly 6.927597e-03
How.Long.Internet.Daily.Hour 3.612047e-03
Energy.efficiency .
Metal -6.556145e-02
Paper -7.009317e-02
Plastic -4.427840e-02
Glass -3.604447e-02
Stove 8.269632e-03
Oven 2.230270e-02
Microwave 3.326151e-04
Grill 1.333591e-02
Airfryer 3.105221e-15zeroCoeff<-coeff==0
zeroCoeff27 x 1 Matrix of class "lgeMatrix"
s0
(Intercept) FALSE
Body.Type TRUE
Sex TRUE
Diet TRUE
How.Often.Shower TRUE
Heating.Energy.Source TRUE
Transport TRUE
Vehicle.Type TRUE
Social.Activity TRUE
Monthly.Grocery.Bill FALSE
Frequency.of.Traveling.by.Air TRUE
Vehicle.Monthly.Distance.Km FALSE
Waste.Bag.Size TRUE
Waste.Bag.Weekly.Count FALSE
How.Long.TV.PC.Daily.Hour FALSE
How.Many.New.Clothes.Monthly FALSE
How.Long.Internet.Daily.Hour FALSE
Energy.efficiency TRUE
Metal FALSE
Paper FALSE
Plastic FALSE
Glass FALSE
Stove FALSE
Oven FALSE
Microwave FALSE
Grill FALSE
Airfryer FALSEplot(lassoModel)
#Ridge Model
set.seed(1)
ridgeModel<-train(CarbonEmission~.,data = carbonTrainData,method="glmnet",trControl= trainControl(method = "cv", number=5), tuneGrid = expand.grid(alpha=0, lambda=10^seq(-3,3,length=100))) Warning: There were missing values in resampled performance measures.ridgeModelglmnet
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
lambda RMSE Rsquared MAE
1.000000e-03 0.1241596 0.9222100 0.08865090
1.149757e-03 0.1241596 0.9222100 0.08865090
1.321941e-03 0.1241596 0.9222100 0.08865090
1.519911e-03 0.1241596 0.9222100 0.08865090
1.747528e-03 0.1241596 0.9222100 0.08865090
2.009233e-03 0.1241596 0.9222100 0.08865090
2.310130e-03 0.1241596 0.9222100 0.08865090
2.656088e-03 0.1241596 0.9222100 0.08865090
3.053856e-03 0.1241596 0.9222100 0.08865090
3.511192e-03 0.1241596 0.9222100 0.08865090
4.037017e-03 0.1241596 0.9222100 0.08865090
4.641589e-03 0.1241596 0.9222100 0.08865090
5.336699e-03 0.1241596 0.9222100 0.08865090
6.135907e-03 0.1241596 0.9222100 0.08865090
7.054802e-03 0.1241596 0.9222100 0.08865090
8.111308e-03 0.1241596 0.9222100 0.08865090
9.326033e-03 0.1241596 0.9222100 0.08865090
1.072267e-02 0.1241596 0.9222100 0.08865090
1.232847e-02 0.1241596 0.9222100 0.08865090
1.417474e-02 0.1241596 0.9222100 0.08865090
1.629751e-02 0.1241596 0.9222100 0.08865090
1.873817e-02 0.1241596 0.9222100 0.08865090
2.154435e-02 0.1241596 0.9222100 0.08865090
2.477076e-02 0.1245591 0.9220121 0.08897062
2.848036e-02 0.1253433 0.9216210 0.08960188
3.274549e-02 0.1263232 0.9211336 0.09040156
3.764936e-02 0.1275324 0.9205351 0.09139279
4.328761e-02 0.1290214 0.9197991 0.09261115
4.977024e-02 0.1308283 0.9189101 0.09408119
5.722368e-02 0.1330158 0.9178334 0.09587911
6.579332e-02 0.1356200 0.9165564 0.09804399
7.564633e-02 0.1387138 0.9150356 0.10063548
8.697490e-02 0.1423210 0.9132671 0.10365698
1.000000e-01 0.1465209 0.9111976 0.10718009
1.149757e-01 0.1513120 0.9088398 0.11124324
1.321941e-01 0.1567765 0.9061288 0.11590149
1.519911e-01 0.1628775 0.9030968 0.12112912
1.747528e-01 0.1696971 0.8996651 0.12698776
2.009233e-01 0.1771493 0.8958995 0.13340871
2.310130e-01 0.1853174 0.8916978 0.14044834
2.656088e-01 0.1940616 0.8871586 0.14797845
3.053856e-01 0.2034668 0.8821560 0.15602687
3.511192e-01 0.2133348 0.8768295 0.16444914
4.037017e-01 0.2237552 0.8710266 0.17329088
4.641589e-01 0.2344738 0.8649369 0.18233732
5.336699e-01 0.2455904 0.8583828 0.19165825
6.135907e-01 0.2568040 0.8516140 0.20102781
7.054802e-01 0.2682319 0.8444305 0.21057248
8.111308e-01 0.2795432 0.8371551 0.22002436
9.326033e-01 0.2908766 0.8295481 0.22946455
1.072267e+00 0.3018907 0.8219941 0.23860783
1.232847e+00 0.3127525 0.8142487 0.24762649
1.417474e+00 0.3231279 0.8067264 0.25623745
1.629751e+00 0.3332082 0.7991576 0.26457118
1.873817e+00 0.3426839 0.7919688 0.27238956
2.154435e+00 0.3517637 0.7848677 0.27986696
2.477076e+00 0.3601737 0.7782639 0.28677239
2.848036e+00 0.3681314 0.7718504 0.29329776
3.274549e+00 0.3754039 0.7659986 0.29926155
3.764936e+00 0.3822074 0.7603999 0.30482829
4.328761e+00 0.3883513 0.7553751 0.30984318
4.977024e+00 0.3940416 0.7506290 0.31448810
5.722368e+00 0.3991267 0.7464272 0.31863716
6.579332e+00 0.4037957 0.7424992 0.32244208
7.564633e+00 0.4079310 0.7390616 0.32580772
8.697490e+00 0.4117002 0.7358752 0.32887573
1.000000e+01 0.4150136 0.7331118 0.33157274
1.149757e+01 0.4180152 0.7305674 0.33401686
1.321941e+01 0.4206378 0.7283766 0.33615249
1.519911e+01 0.4230018 0.7263700 0.33807728
1.747528e+01 0.4250570 0.7246521 0.33974864
2.009233e+01 0.4269022 0.7230850 0.34124719
2.310130e+01 0.4285000 0.7217491 0.34254381
2.656088e+01 0.4299301 0.7205344 0.34370335
3.053856e+01 0.4311646 0.7195024 0.34470425
3.511192e+01 0.4322668 0.7185663 0.34559733
4.037017e+01 0.4332159 0.7177731 0.34636608
4.641589e+01 0.4340617 0.7170549 0.34705090
5.336699e+01 0.4347886 0.7164475 0.34763924
6.135907e+01 0.4354354 0.7158985 0.34816259
7.054802e+01 0.4359905 0.7154352 0.34861180
8.111308e+01 0.4364839 0.7150163 0.34901104
9.326033e+01 0.4369069 0.7146629 0.34935317
1.072267e+02 0.4372824 0.7143440 0.34965704
1.232847e+02 0.4376042 0.7140753 0.34991749
1.417474e+02 0.4378896 0.7138330 0.35014864
1.629751e+02 0.4381340 0.7136289 0.35034645
1.873817e+02 0.4383508 0.7134450 0.35052188
2.154435e+02 0.4390209 0.7133272 0.35106476
2.477076e+02 0.4397931 NaN 0.35168920
2.848036e+02 0.4397931 NaN 0.35168920
3.274549e+02 0.4397931 NaN 0.35168920
3.764936e+02 0.4397931 NaN 0.35168920
4.328761e+02 0.4397931 NaN 0.35168920
4.977024e+02 0.4397931 NaN 0.35168920
5.722368e+02 0.4397931 NaN 0.35168920
6.579332e+02 0.4397931 NaN 0.35168920
7.564633e+02 0.4397931 NaN 0.35168920
8.697490e+02 0.4397931 NaN 0.35168920
1.000000e+03 0.4397931 NaN 0.35168920
Tuning parameter 'alpha' was held constant at a value of 0
RMSE was used to select the optimal model using the smallest value.
The final values used for the model were alpha = 0 and lambda = 0.02154435.ridgeLambda<-ridgeModel$bestTune$lambda
ridgePredictor<- setdiff(names(carbonTrainData),"CarbonEmission")
ridgeFinalModel<-glmnet(as.matrix(carbonTrainData[,ridgePredictor]),carbonTrainData[,"CarbonEmission"],alpha = 1,lambda = ridgeLambda, family = "gaussian")Warning: NAs introduced by coercionridgeFinalModel
Call: glmnet(x = as.matrix(carbonTrainData[, ridgePredictor]), y = carbonTrainData[, "CarbonEmission"], family = "gaussian", alpha = 1, lambda = ridgeLambda)
Df %Dev Lambda
1 8 36.64 0.02154plot(ridgeModel)
set.seed(1)
enetModel<-train(CarbonEmission~., data = carbonTrainData, method = "glmnet", trControl=trainControl(method="cv",number=5,preProc="nzv"),tuneGrid=expand.grid(alpha=seq(0,1,length=10),lambda=10^seq(-3,1,length=100)))Warning: There were missing values in resampled performance measures.enetModelglmnet
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
alpha lambda RMSE Rsquared MAE
0.0000000 0.001000000 0.1241596 0.9222100 0.08865090
0.0000000 0.001097499 0.1241596 0.9222100 0.08865090
0.0000000 0.001204504 0.1241596 0.9222100 0.08865090
0.0000000 0.001321941 0.1241596 0.9222100 0.08865090
0.0000000 0.001450829 0.1241596 0.9222100 0.08865090
0.0000000 0.001592283 0.1241596 0.9222100 0.08865090
0.0000000 0.001747528 0.1241596 0.9222100 0.08865090
0.0000000 0.001917910 0.1241596 0.9222100 0.08865090
0.0000000 0.002104904 0.1241596 0.9222100 0.08865090
0.0000000 0.002310130 0.1241596 0.9222100 0.08865090
0.0000000 0.002535364 0.1241596 0.9222100 0.08865090
0.0000000 0.002782559 0.1241596 0.9222100 0.08865090
0.0000000 0.003053856 0.1241596 0.9222100 0.08865090
0.0000000 0.003351603 0.1241596 0.9222100 0.08865090
0.0000000 0.003678380 0.1241596 0.9222100 0.08865090
0.0000000 0.004037017 0.1241596 0.9222100 0.08865090
0.0000000 0.004430621 0.1241596 0.9222100 0.08865090
0.0000000 0.004862602 0.1241596 0.9222100 0.08865090
0.0000000 0.005336699 0.1241596 0.9222100 0.08865090
0.0000000 0.005857021 0.1241596 0.9222100 0.08865090
0.0000000 0.006428073 0.1241596 0.9222100 0.08865090
0.0000000 0.007054802 0.1241596 0.9222100 0.08865090
0.0000000 0.007742637 0.1241596 0.9222100 0.08865090
0.0000000 0.008497534 0.1241596 0.9222100 0.08865090
0.0000000 0.009326033 0.1241596 0.9222100 0.08865090
0.0000000 0.010235310 0.1241596 0.9222100 0.08865090
0.0000000 0.011233240 0.1241596 0.9222100 0.08865090
0.0000000 0.012328467 0.1241596 0.9222100 0.08865090
0.0000000 0.013530478 0.1241596 0.9222100 0.08865090
0.0000000 0.014849683 0.1241596 0.9222100 0.08865090
0.0000000 0.016297508 0.1241596 0.9222100 0.08865090
0.0000000 0.017886495 0.1241596 0.9222100 0.08865090
0.0000000 0.019630407 0.1241596 0.9222100 0.08865090
0.0000000 0.021544347 0.1241596 0.9222100 0.08865090
0.0000000 0.023644894 0.1243336 0.9221251 0.08878981
0.0000000 0.025950242 0.1248007 0.9218916 0.08916443
0.0000000 0.028480359 0.1253433 0.9216210 0.08960188
0.0000000 0.031257158 0.1259718 0.9213084 0.09011437
0.0000000 0.034304693 0.1266974 0.9209484 0.09070877
0.0000000 0.037649358 0.1275324 0.9205351 0.09139279
0.0000000 0.041320124 0.1284898 0.9200622 0.09217462
0.0000000 0.045348785 0.1295836 0.9195229 0.09307200
0.0000000 0.049770236 0.1308283 0.9189101 0.09408119
0.0000000 0.054622772 0.1322389 0.9182166 0.09523806
0.0000000 0.059948425 0.1338309 0.9174346 0.09655549
0.0000000 0.065793322 0.1356200 0.9165564 0.09804399
0.0000000 0.072208090 0.1376214 0.9155743 0.09971664
0.0000000 0.079248290 0.1398503 0.9144804 0.10158536
0.0000000 0.086974900 0.1423210 0.9132671 0.10365698
0.0000000 0.095454846 0.1450470 0.9119271 0.10593910
0.0000000 0.104761575 0.1480405 0.9104532 0.10846769
0.0000000 0.114975700 0.1513120 0.9088398 0.11124324
0.0000000 0.126185688 0.1548706 0.9070803 0.11427081
0.0000000 0.138488637 0.1587242 0.9051676 0.11756418
0.0000000 0.151991108 0.1628775 0.9030968 0.12112912
0.0000000 0.166810054 0.1673331 0.9008642 0.12496092
0.0000000 0.183073828 0.1720912 0.8984661 0.12905094
0.0000000 0.200923300 0.1771493 0.8958995 0.13340871
0.0000000 0.220513074 0.1825026 0.8931607 0.13802299
0.0000000 0.242012826 0.1881435 0.8902477 0.14288382
0.0000000 0.265608778 0.1940616 0.8871586 0.14797845
0.0000000 0.291505306 0.2002439 0.8838925 0.15327021
0.0000000 0.319926714 0.2066744 0.8804492 0.15876514
0.0000000 0.351119173 0.2133348 0.8768295 0.16444914
0.0000000 0.385352859 0.2202039 0.8730351 0.17028302
0.0000000 0.422924287 0.2272587 0.8690693 0.17625470
0.0000000 0.464158883 0.2344738 0.8649369 0.18233732
0.0000000 0.509413801 0.2418221 0.8606442 0.18850337
0.0000000 0.559081018 0.2492753 0.8561998 0.19473774
0.0000000 0.613590727 0.2568040 0.8516140 0.20102781
0.0000000 0.673415066 0.2643782 0.8468997 0.20735763
0.0000000 0.739072203 0.2719680 0.8420750 0.21369430
0.0000000 0.811130831 0.2795432 0.8371551 0.22002436
0.0000000 0.890215085 0.2870737 0.8321512 0.22630021
0.0000000 0.977009957 0.2945319 0.8270902 0.23250193
0.0000000 1.072267222 0.3018907 0.8219941 0.23860783
0.0000000 1.176811952 0.3091249 0.8168859 0.24461487
0.0000000 1.291549665 0.3162111 0.8117888 0.25049901
0.0000000 1.417474163 0.3231279 0.8067264 0.25623745
0.0000000 1.555676144 0.3298564 0.8017212 0.26180204
0.0000000 1.707352647 0.3363798 0.7967952 0.26719085
0.0000000 1.873817423 0.3426839 0.7919688 0.27238956
0.0000000 2.056512308 0.3487568 0.7872607 0.27739121
0.0000000 2.257019720 0.3545891 0.7826876 0.28218926
0.0000000 2.477076356 0.3601737 0.7782639 0.28677239
0.0000000 2.718588243 0.3655059 0.7740017 0.29114498
0.0000000 2.983647240 0.3705828 0.7699108 0.29530828
0.0000000 3.274549163 0.3754039 0.7659986 0.29926155
0.0000000 3.593813664 0.3799702 0.7622703 0.30299929
0.0000000 3.944206059 0.3842846 0.7587287 0.30652481
0.0000000 4.328761281 0.3883513 0.7553751 0.30984318
0.0000000 4.750810162 0.3921759 0.7522089 0.31296488
0.0000000 5.214008288 0.3957652 0.7492277 0.31589452
0.0000000 5.722367659 0.3991267 0.7464272 0.31863716
0.0000000 6.280291442 0.4022688 0.7438033 0.32119875
0.0000000 6.892612104 0.4052005 0.7413502 0.32358524
0.0000000 7.564633276 0.4079310 0.7390616 0.32580772
0.0000000 8.302175681 0.4104702 0.7369306 0.32787458
0.0000000 9.111627561 0.4128278 0.7349499 0.32979342
0.0000000 10.000000000 0.4150136 0.7331118 0.33157274
0.1111111 0.001000000 0.1217334 0.9234459 0.08677404
0.1111111 0.001097499 0.1217334 0.9234459 0.08677404
0.1111111 0.001204504 0.1217334 0.9234459 0.08677404
0.1111111 0.001321941 0.1217334 0.9234459 0.08677404
0.1111111 0.001450829 0.1217334 0.9234462 0.08677364
0.1111111 0.001592283 0.1217397 0.9234421 0.08677424
0.1111111 0.001747528 0.1217472 0.9234374 0.08677558
0.1111111 0.001917910 0.1217557 0.9234320 0.08677780
0.1111111 0.002104904 0.1217665 0.9234254 0.08678165
0.1111111 0.002310130 0.1217774 0.9234192 0.08678479
0.1111111 0.002535364 0.1217927 0.9234097 0.08679195
0.1111111 0.002782559 0.1218112 0.9233983 0.08680125
0.1111111 0.003053856 0.1218335 0.9233846 0.08681348
0.1111111 0.003351603 0.1218601 0.9233682 0.08682921
0.1111111 0.003678380 0.1218921 0.9233485 0.08684948
0.1111111 0.004037017 0.1219305 0.9233250 0.08687637
0.1111111 0.004430621 0.1219764 0.9232968 0.08690955
0.1111111 0.004862602 0.1220311 0.9232633 0.08694945
0.1111111 0.005336699 0.1220956 0.9232239 0.08699853
0.1111111 0.005857021 0.1221727 0.9231779 0.08705854
0.1111111 0.006428073 0.1222625 0.9231235 0.08712762
0.1111111 0.007054802 0.1223716 0.9230582 0.08721088
0.1111111 0.007742637 0.1225033 0.9229780 0.08731223
0.1111111 0.008497534 0.1226601 0.9228823 0.08743698
0.1111111 0.009326033 0.1228463 0.9227684 0.08758640
0.1111111 0.010235310 0.1230668 0.9226335 0.08776182
0.1111111 0.011233240 0.1233256 0.9224757 0.08796704
0.1111111 0.012328467 0.1236361 0.9222828 0.08821292
0.1111111 0.013530478 0.1239985 0.9220595 0.08850432
0.1111111 0.014849683 0.1244299 0.9217901 0.08885240
0.1111111 0.016297508 0.1249354 0.9214730 0.08927013
0.1111111 0.017886495 0.1255325 0.9210947 0.08977170
0.1111111 0.019630407 0.1262359 0.9206431 0.09036904
0.1111111 0.021544347 0.1270514 0.9201159 0.09106724
0.1111111 0.023644894 0.1280062 0.9194893 0.09188451
0.1111111 0.025950242 0.1291143 0.9187547 0.09282390
0.1111111 0.028480359 0.1303934 0.9178960 0.09391840
0.1111111 0.031257158 0.1318723 0.9168851 0.09519223
0.1111111 0.034304693 0.1335698 0.9157020 0.09664438
0.1111111 0.037649358 0.1355141 0.9143136 0.09831634
0.1111111 0.041320124 0.1377232 0.9127111 0.10017116
0.1111111 0.045348785 0.1402113 0.9108721 0.10224274
0.1111111 0.049770236 0.1430276 0.9087235 0.10459590
0.1111111 0.054622772 0.1461284 0.9063414 0.10718878
0.1111111 0.059948425 0.1494363 0.9038941 0.10994517
0.1111111 0.065793322 0.1529902 0.9013442 0.11292117
0.1111111 0.072208090 0.1567623 0.8987544 0.11610390
0.1111111 0.079248290 0.1608681 0.8959800 0.11954882
0.1111111 0.086974900 0.1653875 0.8928412 0.12334553
0.1111111 0.095454846 0.1703359 0.8893017 0.12748123
0.1111111 0.104761575 0.1756553 0.8854898 0.13194740
0.1111111 0.114975700 0.1812419 0.8816911 0.13661277
0.1111111 0.126185688 0.1872992 0.8774701 0.14171717
0.1111111 0.138488637 0.1938499 0.8727256 0.14722242
0.1111111 0.151991108 0.2009144 0.8673238 0.15314502
0.1111111 0.166810054 0.2084710 0.8611874 0.15946882
0.1111111 0.183073828 0.2165904 0.8540000 0.16627763
0.1111111 0.200923300 0.2253002 0.8454287 0.17358788
0.1111111 0.220513074 0.2344942 0.8355983 0.18126942
0.1111111 0.242012826 0.2438593 0.8255182 0.18905006
0.1111111 0.265608778 0.2535111 0.8147922 0.19704125
0.1111111 0.291505306 0.2634364 0.8032475 0.20522311
0.1111111 0.319926714 0.2737713 0.7898724 0.21375087
0.1111111 0.351119173 0.2842888 0.7753295 0.22246859
0.1111111 0.385352859 0.2949711 0.7595295 0.23134406
0.1111111 0.422924287 0.3060137 0.7405776 0.24047033
0.1111111 0.464158883 0.3174432 0.7170064 0.24990101
0.1111111 0.509413801 0.3291159 0.6882222 0.25951702
0.1111111 0.559081018 0.3405148 0.6571255 0.26895359
0.1111111 0.613590727 0.3517377 0.6222571 0.27825887
0.1111111 0.673415066 0.3618434 0.5940626 0.28668262
0.1111111 0.739072203 0.3710416 0.5715066 0.29438372
0.1111111 0.811130831 0.3793860 0.5563613 0.30128810
0.1111111 0.890215085 0.3874181 0.5409958 0.30792701
0.1111111 0.977009957 0.3950869 0.5281318 0.31427023
0.1111111 1.072267222 0.4024619 0.5182535 0.32037846
0.1111111 1.176811952 0.4097467 0.5038059 0.32646378
0.1111111 1.291549665 0.4165295 0.4890961 0.33224260
0.1111111 1.417474163 0.4231500 0.4611175 0.33789654
0.1111111 1.555676144 0.4288538 0.4430035 0.34277159
0.1111111 1.707352647 0.4341464 0.4012099 0.34727384
0.1111111 1.873817423 0.4377089 0.2710946 0.35022545
0.1111111 2.056512308 0.4397661 0.2502579 0.35167019
0.1111111 2.257019720 0.4397931 NaN 0.35168920
0.1111111 2.477076356 0.4397931 NaN 0.35168920
0.1111111 2.718588243 0.4397931 NaN 0.35168920
0.1111111 2.983647240 0.4397931 NaN 0.35168920
0.1111111 3.274549163 0.4397931 NaN 0.35168920
0.1111111 3.593813664 0.4397931 NaN 0.35168920
0.1111111 3.944206059 0.4397931 NaN 0.35168920
0.1111111 4.328761281 0.4397931 NaN 0.35168920
0.1111111 4.750810162 0.4397931 NaN 0.35168920
0.1111111 5.214008288 0.4397931 NaN 0.35168920
0.1111111 5.722367659 0.4397931 NaN 0.35168920
0.1111111 6.280291442 0.4397931 NaN 0.35168920
0.1111111 6.892612104 0.4397931 NaN 0.35168920
0.1111111 7.564633276 0.4397931 NaN 0.35168920
0.1111111 8.302175681 0.4397931 NaN 0.35168920
0.1111111 9.111627561 0.4397931 NaN 0.35168920
0.1111111 10.000000000 0.4397931 NaN 0.35168920
[ reached getOption("max.print") -- omitted 800 rows ]
RMSE was used to select the optimal model using the smallest value.
The final values used for the model were alpha = 0.1111111 and lambda = 0.001321941.enetModel$bestTuneenetLambda<-enetModel$bestTune$lambda
enetAlpha<-enetModel$bestTune$alpha
enetPredector<-setdiff(names(carbonTrainData),"CarbonEmission")
enetFinalModel<-glmnet(as.matrix(carbonTrainData[,enetPredector]),carbonTrainData[,"CarbonEmission"], alpha = enetAlpha,lambda = enetLambda, family = "gaussian")Warning: NAs introduced by coercionenetFinalModel
Call: glmnet(x = as.matrix(carbonTrainData[, enetPredector]), y = carbonTrainData[, "CarbonEmission"], family = "gaussian", alpha = enetAlpha, lambda = enetLambda)
Df %Dev Lambda
1 15 38.87 0.001322#Random Forest Model
library(randomForest)randomForest 4.7-1.1
Type rfNews() to see new features/changes/bug fixes.
Attaching package: ‘randomForest’
The following object is masked from ‘package:dplyr’:
combine
The following object is masked from ‘package:ggplot2’:
marginset.seed(1)
randomForestModel<-randomForest(CarbonEmission~.,data = carbonTrainData)
randomForestModel
Call:
randomForest(formula = CarbonEmission ~ ., data = carbonTrainData)
Type of random forest: regression
Number of trees: 500
No. of variables tried at each split: 8
Mean of squared residuals: 0.01633518
% Var explained: 91.56mRf<-train(CarbonEmission~.,
data=carbonTrainData,
method="rf",
trControl=trainControl(method = "cv", number =5)
)mRfRandom Forest
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
mtry RMSE Rsquared MAE
2 0.2675949 0.8091212 0.2085785
23 0.1383621 0.9051976 0.1068202
45 0.1421811 0.8973471 0.1099139
RMSE was used to select the optimal model using the smallest value.
The final value used for the model was mtry = 23.varImp(mRf)rf variable importance
only 20 most important variables shown (out of 45)rfPred<-predict(mRf,newdata = carbonTestData)
MAE(carbonTestData$CarbonEmission,rfPred)[1] 0.1045264rmse(carbonTestData$CarbonEmission,rfPred)[1] 0.00120078cor(carbonTestData$CarbonEmission,rfPred)^2[1] 0.9072645plot(carbonTestData$CarbonEmission,rfPred)
GBM
set.seed(1)
grBoostedTree<-train(
CarbonEmission~.,
data = carbonTrainData,
method="gbm",
trControl=trainControl(method = "cv",number = 5)
)Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1840 nan 0.1000 0.0083
2 0.1764 nan 0.1000 0.0078
3 0.1696 nan 0.1000 0.0067
4 0.1630 nan 0.1000 0.0063
5 0.1574 nan 0.1000 0.0058
6 0.1521 nan 0.1000 0.0052
7 0.1475 nan 0.1000 0.0047
8 0.1434 nan 0.1000 0.0039
9 0.1392 nan 0.1000 0.0041
10 0.1357 nan 0.1000 0.0032
20 0.1098 nan 0.1000 0.0017
40 0.0855 nan 0.1000 0.0009
60 0.0696 nan 0.1000 0.0006
80 0.0584 nan 0.1000 0.0005
100 0.0502 nan 0.1000 0.0004
120 0.0441 nan 0.1000 0.0002
140 0.0395 nan 0.1000 0.0002
150 0.0375 nan 0.1000 0.0002
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1765 nan 0.1000 0.0158
2 0.1638 nan 0.1000 0.0126
3 0.1535 nan 0.1000 0.0102
4 0.1447 nan 0.1000 0.0082
5 0.1374 nan 0.1000 0.0074
6 0.1313 nan 0.1000 0.0060
7 0.1249 nan 0.1000 0.0061
8 0.1198 nan 0.1000 0.0053
9 0.1148 nan 0.1000 0.0049
10 0.1105 nan 0.1000 0.0043
20 0.0804 nan 0.1000 0.0028
40 0.0534 nan 0.1000 0.0009
60 0.0394 nan 0.1000 0.0004
80 0.0308 nan 0.1000 0.0003
100 0.0256 nan 0.1000 0.0002
120 0.0215 nan 0.1000 0.0001
140 0.0187 nan 0.1000 0.0001
150 0.0176 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1740 nan 0.1000 0.0187
2 0.1583 nan 0.1000 0.0156
3 0.1457 nan 0.1000 0.0125
4 0.1349 nan 0.1000 0.0105
5 0.1264 nan 0.1000 0.0083
6 0.1184 nan 0.1000 0.0079
7 0.1119 nan 0.1000 0.0062
8 0.1057 nan 0.1000 0.0059
9 0.1008 nan 0.1000 0.0049
10 0.0961 nan 0.1000 0.0044
20 0.0673 nan 0.1000 0.0019
40 0.0412 nan 0.1000 0.0010
60 0.0290 nan 0.1000 0.0004
80 0.0224 nan 0.1000 0.0003
100 0.0179 nan 0.1000 0.0001
120 0.0146 nan 0.1000 0.0001
140 0.0125 nan 0.1000 0.0001
150 0.0116 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1856 nan 0.1000 0.0089
2 0.1777 nan 0.1000 0.0076
3 0.1703 nan 0.1000 0.0073
4 0.1641 nan 0.1000 0.0062
5 0.1583 nan 0.1000 0.0058
6 0.1533 nan 0.1000 0.0048
7 0.1482 nan 0.1000 0.0050
8 0.1441 nan 0.1000 0.0040
9 0.1404 nan 0.1000 0.0037
10 0.1366 nan 0.1000 0.0039
20 0.1116 nan 0.1000 0.0017
40 0.0869 nan 0.1000 0.0008
60 0.0712 nan 0.1000 0.0006
80 0.0598 nan 0.1000 0.0004
100 0.0514 nan 0.1000 0.0003
120 0.0451 nan 0.1000 0.0002
140 0.0405 nan 0.1000 0.0002
150 0.0384 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1785 nan 0.1000 0.0156
2 0.1654 nan 0.1000 0.0126
3 0.1548 nan 0.1000 0.0105
4 0.1459 nan 0.1000 0.0086
5 0.1386 nan 0.1000 0.0071
6 0.1323 nan 0.1000 0.0060
7 0.1271 nan 0.1000 0.0052
8 0.1219 nan 0.1000 0.0050
9 0.1161 nan 0.1000 0.0058
10 0.1117 nan 0.1000 0.0043
20 0.0824 nan 0.1000 0.0016
40 0.0548 nan 0.1000 0.0007
60 0.0400 nan 0.1000 0.0005
80 0.0308 nan 0.1000 0.0004
100 0.0254 nan 0.1000 0.0002
120 0.0213 nan 0.1000 0.0001
140 0.0185 nan 0.1000 0.0001
150 0.0174 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1760 nan 0.1000 0.0187
2 0.1605 nan 0.1000 0.0159
3 0.1476 nan 0.1000 0.0124
4 0.1376 nan 0.1000 0.0102
5 0.1286 nan 0.1000 0.0088
6 0.1201 nan 0.1000 0.0081
7 0.1133 nan 0.1000 0.0067
8 0.1070 nan 0.1000 0.0064
9 0.1019 nan 0.1000 0.0050
10 0.0971 nan 0.1000 0.0048
20 0.0689 nan 0.1000 0.0024
40 0.0422 nan 0.1000 0.0007
60 0.0300 nan 0.1000 0.0006
80 0.0227 nan 0.1000 0.0002
100 0.0181 nan 0.1000 0.0002
120 0.0148 nan 0.1000 0.0001
140 0.0126 nan 0.1000 0.0001
150 0.0117 nan 0.1000 0.0000
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1849 nan 0.1000 0.0087
2 0.1772 nan 0.1000 0.0079
3 0.1697 nan 0.1000 0.0072
4 0.1635 nan 0.1000 0.0062
5 0.1574 nan 0.1000 0.0058
6 0.1524 nan 0.1000 0.0046
7 0.1472 nan 0.1000 0.0051
8 0.1430 nan 0.1000 0.0040
9 0.1389 nan 0.1000 0.0043
10 0.1352 nan 0.1000 0.0035
20 0.1105 nan 0.1000 0.0016
40 0.0859 nan 0.1000 0.0010
60 0.0702 nan 0.1000 0.0006
80 0.0586 nan 0.1000 0.0004
100 0.0504 nan 0.1000 0.0003
120 0.0440 nan 0.1000 0.0002
140 0.0394 nan 0.1000 0.0001
150 0.0374 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1777 nan 0.1000 0.0156
2 0.1646 nan 0.1000 0.0131
3 0.1540 nan 0.1000 0.0103
4 0.1448 nan 0.1000 0.0089
5 0.1373 nan 0.1000 0.0073
6 0.1309 nan 0.1000 0.0061
7 0.1245 nan 0.1000 0.0064
8 0.1194 nan 0.1000 0.0049
9 0.1142 nan 0.1000 0.0053
10 0.1102 nan 0.1000 0.0040
20 0.0812 nan 0.1000 0.0018
40 0.0536 nan 0.1000 0.0010
60 0.0393 nan 0.1000 0.0005
80 0.0308 nan 0.1000 0.0004
100 0.0252 nan 0.1000 0.0001
120 0.0213 nan 0.1000 0.0002
140 0.0186 nan 0.1000 0.0001
150 0.0175 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1741 nan 0.1000 0.0187
2 0.1585 nan 0.1000 0.0155
3 0.1461 nan 0.1000 0.0126
4 0.1357 nan 0.1000 0.0103
5 0.1255 nan 0.1000 0.0100
6 0.1181 nan 0.1000 0.0075
7 0.1111 nan 0.1000 0.0071
8 0.1053 nan 0.1000 0.0056
9 0.1002 nan 0.1000 0.0050
10 0.0958 nan 0.1000 0.0045
20 0.0675 nan 0.1000 0.0020
40 0.0421 nan 0.1000 0.0007
60 0.0296 nan 0.1000 0.0003
80 0.0223 nan 0.1000 0.0003
100 0.0181 nan 0.1000 0.0001
120 0.0150 nan 0.1000 0.0001
140 0.0126 nan 0.1000 0.0001
150 0.0114 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1851 nan 0.1000 0.0086
2 0.1775 nan 0.1000 0.0078
3 0.1701 nan 0.1000 0.0073
4 0.1635 nan 0.1000 0.0064
5 0.1578 nan 0.1000 0.0057
6 0.1527 nan 0.1000 0.0049
7 0.1476 nan 0.1000 0.0049
8 0.1436 nan 0.1000 0.0039
9 0.1396 nan 0.1000 0.0039
10 0.1356 nan 0.1000 0.0038
20 0.1116 nan 0.1000 0.0018
40 0.0869 nan 0.1000 0.0009
60 0.0711 nan 0.1000 0.0005
80 0.0596 nan 0.1000 0.0005
100 0.0513 nan 0.1000 0.0003
120 0.0452 nan 0.1000 0.0002
140 0.0404 nan 0.1000 0.0001
150 0.0384 nan 0.1000 0.0002
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1785 nan 0.1000 0.0155
2 0.1654 nan 0.1000 0.0128
3 0.1552 nan 0.1000 0.0102
4 0.1464 nan 0.1000 0.0088
5 0.1388 nan 0.1000 0.0076
6 0.1324 nan 0.1000 0.0063
7 0.1262 nan 0.1000 0.0064
8 0.1208 nan 0.1000 0.0052
9 0.1156 nan 0.1000 0.0053
10 0.1110 nan 0.1000 0.0045
20 0.0824 nan 0.1000 0.0016
40 0.0554 nan 0.1000 0.0008
60 0.0406 nan 0.1000 0.0005
80 0.0321 nan 0.1000 0.0002
100 0.0265 nan 0.1000 0.0003
120 0.0222 nan 0.1000 0.0002
140 0.0193 nan 0.1000 0.0001
150 0.0180 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1748 nan 0.1000 0.0191
2 0.1595 nan 0.1000 0.0156
3 0.1467 nan 0.1000 0.0124
4 0.1366 nan 0.1000 0.0097
5 0.1272 nan 0.1000 0.0092
6 0.1187 nan 0.1000 0.0084
7 0.1122 nan 0.1000 0.0063
8 0.1061 nan 0.1000 0.0061
9 0.1011 nan 0.1000 0.0050
10 0.0968 nan 0.1000 0.0040
20 0.0688 nan 0.1000 0.0018
40 0.0424 nan 0.1000 0.0007
60 0.0301 nan 0.1000 0.0003
80 0.0229 nan 0.1000 0.0002
100 0.0185 nan 0.1000 0.0002
120 0.0151 nan 0.1000 0.0002
140 0.0129 nan 0.1000 0.0001
150 0.0118 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1834 nan 0.1000 0.0087
2 0.1759 nan 0.1000 0.0075
3 0.1685 nan 0.1000 0.0070
4 0.1627 nan 0.1000 0.0056
5 0.1568 nan 0.1000 0.0062
6 0.1518 nan 0.1000 0.0048
7 0.1467 nan 0.1000 0.0049
8 0.1426 nan 0.1000 0.0039
9 0.1384 nan 0.1000 0.0042
10 0.1348 nan 0.1000 0.0038
20 0.1101 nan 0.1000 0.0018
40 0.0860 nan 0.1000 0.0010
60 0.0703 nan 0.1000 0.0006
80 0.0589 nan 0.1000 0.0004
100 0.0506 nan 0.1000 0.0003
120 0.0445 nan 0.1000 0.0002
140 0.0398 nan 0.1000 0.0002
150 0.0379 nan 0.1000 0.0002
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1763 nan 0.1000 0.0158
2 0.1632 nan 0.1000 0.0127
3 0.1532 nan 0.1000 0.0101
4 0.1441 nan 0.1000 0.0085
5 0.1368 nan 0.1000 0.0071
6 0.1307 nan 0.1000 0.0063
7 0.1249 nan 0.1000 0.0058
8 0.1195 nan 0.1000 0.0055
9 0.1148 nan 0.1000 0.0046
10 0.1102 nan 0.1000 0.0045
20 0.0808 nan 0.1000 0.0018
40 0.0540 nan 0.1000 0.0009
60 0.0400 nan 0.1000 0.0004
80 0.0315 nan 0.1000 0.0003
100 0.0258 nan 0.1000 0.0002
120 0.0217 nan 0.1000 0.0001
140 0.0189 nan 0.1000 0.0001
150 0.0178 nan 0.1000 0.0001
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1734 nan 0.1000 0.0186
2 0.1582 nan 0.1000 0.0152
3 0.1455 nan 0.1000 0.0127
4 0.1352 nan 0.1000 0.0103
5 0.1257 nan 0.1000 0.0094
6 0.1173 nan 0.1000 0.0082
7 0.1109 nan 0.1000 0.0063
8 0.1051 nan 0.1000 0.0056
9 0.1000 nan 0.1000 0.0052
10 0.0952 nan 0.1000 0.0047
20 0.0680 nan 0.1000 0.0018
40 0.0416 nan 0.1000 0.0009
60 0.0290 nan 0.1000 0.0004
80 0.0218 nan 0.1000 0.0002
100 0.0174 nan 0.1000 0.0001
120 0.0141 nan 0.1000 0.0003
140 0.0120 nan 0.1000 0.0001
150 0.0111 nan 0.1000 0.0000
Iter TrainDeviance ValidDeviance StepSize Improve
1 0.1743 nan 0.1000 0.0190
2 0.1589 nan 0.1000 0.0156
3 0.1461 nan 0.1000 0.0125
4 0.1359 nan 0.1000 0.0102
5 0.1262 nan 0.1000 0.0095
6 0.1189 nan 0.1000 0.0073
7 0.1115 nan 0.1000 0.0073
8 0.1056 nan 0.1000 0.0056
9 0.1005 nan 0.1000 0.0051
10 0.0959 nan 0.1000 0.0046
20 0.0678 nan 0.1000 0.0018
40 0.0417 nan 0.1000 0.0007
60 0.0294 nan 0.1000 0.0003
80 0.0224 nan 0.1000 0.0003
100 0.0179 nan 0.1000 0.0001
120 0.0148 nan 0.1000 0.0001
140 0.0125 nan 0.1000 0.0000
150 0.0116 nan 0.1000 0.0001grBoostedTreeStochastic Gradient Boosting
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
interaction.depth n.trees RMSE Rsquared MAE
1 50 0.2799859 0.6817099 0.21734673
1 100 0.2268772 0.7921416 0.17361732
1 150 0.1962693 0.8366385 0.14863782
2 50 0.2171188 0.8190082 0.16598681
2 100 0.1638148 0.8868431 0.12368291
2 150 0.1371742 0.9154801 0.10233794
3 50 0.1894213 0.8550887 0.14434186
3 100 0.1391537 0.9160350 0.10460569
3 150 0.1130848 0.9406167 0.08385162
Tuning parameter 'shrinkage' was held constant at a value of 0.1
Tuning parameter 'n.minobsinnode' was held constant at a value of 10
RMSE was used to select the optimal model using the smallest value.
The final values used for the model were n.trees = 150, interaction.depth = 3, shrinkage = 0.1 and n.minobsinnode = 10.gbmPred<-predict(grBoostedTree, carbonTestData)MAE(carbonTestData$CarbonEmission,gbmPred)[1] 0.08074864rmse(carbonTestData$CarbonEmission,gbmPred)[1] 0.0002656894cor(carbonTestData$CarbonEmission,gbmPred)^2[1] 0.9431872plot(carbonTestData$CarbonEmission,gbmPred)
#SV Linear Model
set.seed(1)
svmLinear<-train(
CarbonEmission~.,
data = carbonTrainData,
method="svmLinear",
trControl=trainControl(method = "cv",number = 5, preProc=c("center","scale"))
)svmLinearSupport Vector Machines with Linear Kernel
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results:
RMSE Rsquared MAE
0.1223375 0.9230945 0.08612926
Tuning parameter 'C' was held constant at a value of 1svmPred<-predict(svmLinear,carbonTestData)
plot(svmPred,carbonTestData$CarbonEmission)
#SVM Radial Model
set.seed(1)
svmRadial<-train(
CarbonEmission~.,
data = carbonTrainData,
method="svmRadial",
trControl=trainControl(method = "cv",number = 5, preProc=c("center","scale"))
)svmRadialSupport Vector Machines with Radial Basis Function Kernel
8001 samples
26 predictor
No pre-processing
Resampling: Cross-Validated (5 fold)
Summary of sample sizes: 6401, 6401, 6401, 6400, 6401
Resampling results across tuning parameters:
C RMSE Rsquared MAE
0.25 0.08070183 0.9684766 0.05543867
0.50 0.06813583 0.9769891 0.04773321
1.00 0.06065717 0.9813954 0.04355294
Tuning parameter 'sigma' was held constant at a value of 0.01193687
RMSE was used to select the optimal model using the smallest value.
The final values used for the model were sigma = 0.01193687 and C = 1.svmRadialPred<-predict(svmRadial,carbonTestData)
plot(svmRadialPred,carbonTestData$CarbonEmission)
#Comparing models
compare=resamples(list(KNN=knnModel,LIN=lmModel,stepWise=stepwiseModel,Lasso=lassoModel,Ridge=ridgeModel,Enet=enetModel,RF=mRf,GBM=grBoostedTree,SVML=svmLinear,SVMR=svmRadial))
summary(compare) # Out of all the models SVM Radial stands out the most
Call:
summary.resamples(object = compare)
Models: KNN, LIN, stepWise, Lasso, Ridge, Enet, RF, GBM, SVML, SVMR
Number of resamples: 5
MAE
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
KNN 0.29355718 0.30011897 0.30358891 0.30104742 0.30359972 0.30437235 0
LIN 0.08468189 0.08653726 0.08668226 0.08674322 0.08771081 0.08810388 0
stepWise 0.27076896 0.27236187 0.27267260 0.27275357 0.27295504 0.27500935 0
Lasso 0.08263733 0.08452922 0.08689541 0.08693488 0.08991372 0.09069875 0
Ridge 0.08461479 0.08660820 0.08863896 0.08865090 0.09132776 0.09206482 0
Enet 0.08278867 0.08412890 0.08658334 0.08677404 0.08993674 0.09043257 0
RF 0.10408932 0.10538108 0.10750605 0.10682021 0.10836087 0.10876372 0
GBM 0.07970661 0.08311406 0.08431983 0.08385162 0.08521232 0.08690528 0
SVML 0.08208710 0.08349344 0.08672664 0.08612926 0.08894605 0.08939306 0
SVMR 0.04203058 0.04268964 0.04371276 0.04355294 0.04405078 0.04528092 0
RMSE
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
KNN 0.37295011 0.37736743 0.37851215 0.37897650 0.37913018 0.3869226 0
LIN 0.11743009 0.12058876 0.12318123 0.12169555 0.12340748 0.1238702 0
stepWise 0.34358303 0.34484586 0.34496809 0.34607902 0.34713797 0.3498602 0
Lasso 0.11662100 0.11703960 0.12364901 0.12198581 0.12624277 0.1263767 0
Ridge 0.11907983 0.11994269 0.12566449 0.12415964 0.12788642 0.1282248 0
Enet 0.11604668 0.11694091 0.12328642 0.12173339 0.12617744 0.1262155 0
RF 0.13408157 0.13614576 0.13894945 0.13836211 0.14047841 0.1421553 0
GBM 0.10685459 0.11180062 0.11338918 0.11308477 0.11588839 0.1174911 0
SVML 0.11628786 0.11705824 0.12519442 0.12233751 0.12626695 0.1268801 0
SVMR 0.05776811 0.05964871 0.06149819 0.06065717 0.06175535 0.0626155 0
Rsquared
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
KNN 0.2543177 0.2631183 0.2650880 0.2674213 0.2767798 0.2778026 0
LIN 0.9185705 0.9216487 0.9246806 0.9235284 0.9253275 0.9274147 0
stepWise 0.3614847 0.3715791 0.3780496 0.3807424 0.3938578 0.3987405 0
Lasso 0.9170290 0.9192558 0.9230649 0.9232209 0.9279729 0.9287817 0
Ridge 0.9155089 0.9187241 0.9226270 0.9222100 0.9266525 0.9275377 0
Enet 0.9173033 0.9192842 0.9233576 0.9234459 0.9285877 0.9286968 0
RF 0.9002318 0.9022657 0.9053106 0.9051976 0.9090824 0.9090975 0
GBM 0.9376708 0.9388972 0.9390649 0.9406167 0.9404415 0.9470088 0
SVML 0.9174402 0.9185159 0.9219757 0.9230945 0.9287006 0.9288401 0
SVMR 0.9805681 0.9808349 0.9812240 0.9813954 0.9815897 0.9827600 0Neural Network Preprocessing
library(caret)
carbonInd<-createDataPartition(carbonTrainData$CarbonEmission,p=0.9,list = FALSE)
carbonIndex<-which(names(carbonTrainData)=='CarbonEmission')
carbonTrainingData<-carbonTrainData[carbonInd,-carbonIndex]
str(carbonTrainingData)'data.frame': 7202 obs. of 26 variables:
$ Body.Type : Factor w/ 4 levels "normal","obese",..: 2 3 2 3 3 4 1 2 4 4 ...
$ Sex : Factor w/ 2 levels "female","male": 1 2 1 2 2 1 1 2 1 1 ...
$ Diet : Factor w/ 4 levels "omnivore","pescatarian",..: 4 1 4 4 1 2 4 4 1 3 ...
$ How.Often.Shower : Factor w/ 4 levels "daily","less frequently",..: 2 3 1 2 1 1 3 3 4 2 ...
$ Heating.Energy.Source : Factor w/ 4 levels "coal","electricity",..: 3 4 1 4 4 4 4 1 1 2 ...
$ Transport : Factor w/ 3 levels "private","public",..: 3 1 1 2 2 2 2 3 3 1 ...
$ Vehicle.Type : Factor w/ 6 levels "diesel","electric",..: 3 6 1 3 3 3 3 3 3 5 ...
$ Social.Activity : Factor w/ 3 levels "never","often",..: 2 1 2 3 1 2 1 1 2 3 ...
$ Monthly.Grocery.Bill : num 114 138 266 144 200 135 146 111 114 111 ...
$ Frequency.of.Traveling.by.Air: Factor w/ 4 levels "frequently","never",..: 3 2 4 1 1 3 2 4 3 3 ...
$ Vehicle.Monthly.Distance.Km : num 9 2472 8457 658 1376 ...
$ Waste.Bag.Size : Factor w/ 4 levels "extra large",..: 1 4 2 2 3 1 1 3 2 2 ...
$ Waste.Bag.Weekly.Count : num 3 1 1 1 3 1 4 5 3 6 ...
$ How.Long.TV.PC.Daily.Hour : num 9 14 3 22 3 8 12 9 18 13 ...
$ How.Many.New.Clothes.Monthly : num 38 47 5 18 31 23 27 4 27 16 ...
$ How.Long.Internet.Daily.Hour : num 5 6 6 9 15 18 21 4 4 10 ...
$ Energy.efficiency : Factor w/ 3 levels "No","Sometimes",..: 1 2 3 2 3 2 1 2 3 2 ...
$ Metal : num 1 1 0 1 0 0 0 0 0 1 ...
$ Paper : num 0 0 1 1 0 0 1 0 0 0 ...
$ Plastic : num 0 0 0 0 0 0 1 0 1 1 ...
$ Glass : num 0 0 0 1 1 1 0 0 0 1 ...
$ Stove : num 1 0 0 1 0 0 1 1 1 1 ...
$ Oven : num 0 1 1 1 0 0 0 1 0 1 ...
$ Microwave : num 1 1 0 1 1 1 1 1 0 1 ...
$ Grill : num 0 0 0 0 1 1 0 0 0 1 ...
$ Airfryer : num 0 0 0 0 1 1 0 0 0 1 ...carbonTrainingLabels<-carbonTrainData[carbonInd,carbonIndex]
str(carbonTrainingLabels) num [1:7202] 7.55 7.86 8.46 7.41 7.82 ...carbonValidationData<-carbonTrainData[-carbonInd,-carbonIndex]
carbonValidationData
carbonValidationLabels<-carbonTrainData[-carbonInd,carbonIndex]
str(carbonValidationLabels) num [1:799] 7.71 7.75 7.09 7.41 7.51 ...carbonTestingData<-carbonTestData[,-carbonIndex]
carbonTestingData
carbonTestingLabels<-carbonTestData[,carbonIndex]
str(carbonTestingLabels) num [1:1999] 6.98 7.51 7.11 7.31 7.48 ...dim(carbonTrainingData)[1] 7202 26dim(carbonTestingData)[1] 1999 26#Scaling numeric Variables and one hot encoding categorical variables
library(mltools)
Attaching package: ‘mltools’
The following object is masked _by_ ‘.GlobalEnv’:
rmse
The following object is masked from ‘package:tidyr’:
replace_nalibrary(data.table)data.table 1.15.4 using 1 threads (see ?getDTthreads). Latest news: r-datatable.com
**********
This installation of data.table has not detected OpenMP support. It should still work but in single-threaded mode.
This is a Mac. Please read https://mac.r-project.org/openmp/. Please engage with Apple and ask them for support. Check r-datatable.com for updates, and our Mac instructions here: https://github.com/Rdatatable/data.table/wiki/Installation. After several years of many reports of installation problems on Mac, it's time to gingerly point out that there have been no similar problems on Windows or Linux.
**********
Attaching package: ‘data.table’
The following objects are masked from ‘package:lubridate’:
hour, isoweek, mday, minute, month, quarter, second, wday, week, yday, year
The following objects are masked from ‘package:dplyr’:
between, first, last
The following object is masked from ‘package:purrr’:
transposenumericCols<-c("Monthly.Grocery.Bill","Vehicle.Monthly.Distance.Km","Waste.Bag.Weekly.Count",
"How.Long.TV.PC.Daily.Hour","How.Many.New.Clothes.Monthly","How.Long.Internet.Daily.Hour","Metal","Paper","Plastic","Glass","Stove","Oven"
,"Microwave","Grill","Airfryer")
categoricalCols<-c("Body.Type","Sex","Diet","How.Often.Shower","Heating.Energy.Source","Transport","Vehicle.Type","Social.Activity",
"Frequency.of.Traveling.by.Air","Waste.Bag.Size","Energy.efficiency")
carbonTrainingDataNew<-scale(carbonTrainingData[,numericCols])
colMeanTrain<-attr(carbonTrainingDataNew,"scaled:center")
colStddevsTrain<-attr(carbonTrainingDataNew,"scaled:scale")
carbonTrainingData[,numericCols]<-carbonTrainingDataNew
carbonValidationData[,numericCols]<-scale(carbonValidationData[,numericCols],center = colMeanTrain,scale = colStddevsTrain)
carbonTestingData[,numericCols]<-scale(carbonTestingData[,numericCols],center = colMeanTrain,scale = colStddevsTrain)
carbonTrainingTable<-as.data.table(carbonTrainingData)
carbonValidationTable<-as.data.table(carbonValidationData)
carbonTestingTable<-as.data.table(carbonTestingData)
carbonTrainingOneHot<-one_hot(carbonTrainingTable,naCols=FALSE,dropCols=TRUE,dropUnusedLevels=TRUE)
carbonTrainingOneHot
carbonValidationOneHot<-one_hot(carbonValidationTable,naCols=FALSE,dropCols=TRUE,dropUnusedLevels=TRUE)
carbonValidationOneHot
carbonTestingOneHot<-one_hot(carbonTestingTable,naCols=FALSE,dropCols=TRUE,dropUnusedLevels=TRUE)
carbonTestingOneHot
carbonTrainingFinal<-as.data.frame(cbind(carbonTrainingTable[, ..numericCols], carbonTrainingOneHot))
carbonTrainingFinal
carbonValidationFinal<-as.data.frame(cbind(carbonValidationTable[, ..numericCols], carbonValidationOneHot))
carbonValidationFinal
carbonTestingFinal<-as.data.frame(cbind(carbonTestingTable[, ..numericCols], carbonTestingOneHot))
carbonTestingFinallibrary(keras)
model<-keras_model_sequential()%>%
layer_dense(units = 32,activation = "relu",input_shape = dim(carbonTrainingFinal)[2])%>%
layer_dropout(rate=0.3)%>%
layer_dense(units = 32,activation = "relu")%>%
layer_dropout(rate=0.3)%>%
layer_dense(units = 16,activation = "relu")%>%
layer_dropout(rate=0.3)%>%
layer_dense(units = 1)
model %>% compile(
loss="mse",
optimizer=optimizer_adam(lr=0.001)
)WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.history<-model %>% fit(as.matrix(carbonTrainingFinal),
carbonTrainingLabels,
batch_size=50,
epochs=20,
validation_data=list(as.matrix(carbonValidationFinal),carbonValidationLabels)
)Epoch 1/202024-05-06 13:56:04.336998: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
1/176 [..............................] - ETA: 1:26 - loss: 52.2808
12/176 [=>............................] - ETA: 0s - loss: 50.3873
25/176 [===>..........................] - ETA: 0s - loss: 43.7593
36/176 [=====>........................] - ETA: 0s - loss: 37.9999
47/176 [=======>......................] - ETA: 0s - loss: 32.8647
60/176 [=========>....................] - ETA: 0s - loss: 28.5552
72/176 [===========>..................] - ETA: 0s - loss: 25.7727
84/176 [=============>................] - ETA: 0s - loss: 23.4563
96/176 [===============>..............] - ETA: 0s - loss: 21.7449
109/176 [=================>............] - ETA: 0s - loss: 20.1764
122/176 [===================>..........] - ETA: 0s - loss: 18.9593
135/176 [======================>.......] - ETA: 0s - loss: 17.9751
149/176 [========================>.....] - ETA: 0s - loss: 16.9692
163/176 [==========================>...] - ETA: 0s - loss: 16.1185
176/176 [==============================] - 1s 4ms/step - loss: 15.45312024-05-06 13:56:05.394796: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
176/176 [==============================] - 2s 8ms/step - loss: 15.4531 - val_loss: 0.9300
Epoch 2/20
1/176 [..............................] - ETA: 1s - loss: 4.9675
13/176 [=>............................] - ETA: 0s - loss: 6.5189
26/176 [===>..........................] - ETA: 0s - loss: 6.1650
40/176 [=====>........................] - ETA: 0s - loss: 6.0132
54/176 [========>.....................] - ETA: 0s - loss: 5.8761
66/176 [==========>...................] - ETA: 0s - loss: 5.8140
80/176 [============>.................] - ETA: 0s - loss: 5.6671
93/176 [==============>...............] - ETA: 0s - loss: 5.5530
106/176 [=================>............] - ETA: 0s - loss: 5.4518
118/176 [===================>..........] - ETA: 0s - loss: 5.3660
132/176 [=====================>........] - ETA: 0s - loss: 5.2230
146/176 [=======================>......] - ETA: 0s - loss: 5.0979
159/176 [==========================>...] - ETA: 0s - loss: 4.9965
172/176 [============================>.] - ETA: 0s - loss: 4.9159
176/176 [==============================] - 1s 4ms/step - loss: 4.8829
176/176 [==============================] - 1s 5ms/step - loss: 4.8829 - val_loss: 0.5066
Epoch 3/20
1/176 [..............................] - ETA: 0s - loss: 3.2023
12/176 [=>............................] - ETA: 0s - loss: 3.5175
25/176 [===>..........................] - ETA: 0s - loss: 3.3779
37/176 [=====>........................] - ETA: 0s - loss: 3.3078
50/176 [=======>......................] - ETA: 0s - loss: 3.2538
63/176 [=========>....................] - ETA: 0s - loss: 3.1842
76/176 [===========>..................] - ETA: 0s - loss: 3.1383
89/176 [==============>...............] - ETA: 0s - loss: 3.0631
102/176 [================>.............] - ETA: 0s - loss: 3.0209
115/176 [==================>...........] - ETA: 0s - loss: 2.9442
129/176 [====================>.........] - ETA: 0s - loss: 2.8997
142/176 [=======================>......] - ETA: 0s - loss: 2.8718
155/176 [=========================>....] - ETA: 0s - loss: 2.8252
169/176 [===========================>..] - ETA: 0s - loss: 2.7705
176/176 [==============================] - 1s 4ms/step - loss: 2.7408
176/176 [==============================] - 1s 5ms/step - loss: 2.7408 - val_loss: 0.3616
Epoch 4/20
1/176 [..............................] - ETA: 0s - loss: 2.5187
13/176 [=>............................] - ETA: 0s - loss: 2.1839
26/176 [===>..........................] - ETA: 0s - loss: 2.1127
39/176 [=====>........................] - ETA: 0s - loss: 2.0266
52/176 [=======>......................] - ETA: 0s - loss: 1.9595
66/176 [==========>...................] - ETA: 0s - loss: 1.9261
79/176 [============>.................] - ETA: 0s - loss: 1.9046
92/176 [==============>...............] - ETA: 0s - loss: 1.8712
105/176 [================>.............] - ETA: 0s - loss: 1.8549
118/176 [===================>..........] - ETA: 0s - loss: 1.8211
131/176 [=====================>........] - ETA: 0s - loss: 1.7954
145/176 [=======================>......] - ETA: 0s - loss: 1.7636
159/176 [==========================>...] - ETA: 0s - loss: 1.7354
173/176 [============================>.] - ETA: 0s - loss: 1.7107
176/176 [==============================] - 1s 4ms/step - loss: 1.7077
176/176 [==============================] - 1s 5ms/step - loss: 1.7077 - val_loss: 0.3143
Epoch 5/20
1/176 [..............................] - ETA: 1s - loss: 1.5128
13/176 [=>............................] - ETA: 0s - loss: 1.3687
26/176 [===>..........................] - ETA: 0s - loss: 1.3742
39/176 [=====>........................] - ETA: 0s - loss: 1.3305
52/176 [=======>......................] - ETA: 0s - loss: 1.3127
66/176 [==========>...................] - ETA: 0s - loss: 1.2746
80/176 [============>.................] - ETA: 0s - loss: 1.2533
91/176 [==============>...............] - ETA: 0s - loss: 1.2272
104/176 [================>.............] - ETA: 0s - loss: 1.2122
117/176 [==================>...........] - ETA: 0s - loss: 1.1916
130/176 [=====================>........] - ETA: 0s - loss: 1.1873
143/176 [=======================>......] - ETA: 0s - loss: 1.1709
156/176 [=========================>....] - ETA: 0s - loss: 1.1531
169/176 [===========================>..] - ETA: 0s - loss: 1.1301
176/176 [==============================] - 1s 4ms/step - loss: 1.1201
176/176 [==============================] - 1s 5ms/step - loss: 1.1201 - val_loss: 0.2966
Epoch 6/20
1/176 [..............................] - ETA: 0s - loss: 0.9772
13/176 [=>............................] - ETA: 0s - loss: 0.9575
21/176 [==>...........................] - ETA: 0s - loss: 0.9310
32/176 [====>.........................] - ETA: 0s - loss: 0.9097
44/176 [======>.......................] - ETA: 0s - loss: 0.9083
57/176 [========>.....................] - ETA: 0s - loss: 0.8997
71/176 [===========>..................] - ETA: 0s - loss: 0.8913
84/176 [=============>................] - ETA: 0s - loss: 0.8834
97/176 [===============>..............] - ETA: 0s - loss: 0.8692
110/176 [=================>............] - ETA: 0s - loss: 0.8619
123/176 [===================>..........] - ETA: 0s - loss: 0.8454
137/176 [======================>.......] - ETA: 0s - loss: 0.8341
150/176 [========================>.....] - ETA: 0s - loss: 0.8313
163/176 [==========================>...] - ETA: 0s - loss: 0.8230
176/176 [==============================] - 1s 4ms/step - loss: 0.8114
176/176 [==============================] - 1s 5ms/step - loss: 0.8114 - val_loss: 0.3290
Epoch 7/20
1/176 [..............................] - ETA: 0s - loss: 1.0693
13/176 [=>............................] - ETA: 0s - loss: 0.8001
26/176 [===>..........................] - ETA: 0s - loss: 0.7705
39/176 [=====>........................] - ETA: 0s - loss: 0.7201
52/176 [=======>......................] - ETA: 0s - loss: 0.7061
65/176 [==========>...................] - ETA: 0s - loss: 0.6947
78/176 [============>.................] - ETA: 0s - loss: 0.6777
91/176 [==============>...............] - ETA: 0s - loss: 0.6651
102/176 [================>.............] - ETA: 0s - loss: 0.6578
114/176 [==================>...........] - ETA: 0s - loss: 0.6578
126/176 [====================>.........] - ETA: 0s - loss: 0.6539
139/176 [======================>.......] - ETA: 0s - loss: 0.6472
152/176 [========================>.....] - ETA: 0s - loss: 0.6426
166/176 [===========================>..] - ETA: 0s - loss: 0.6386
176/176 [==============================] - 1s 4ms/step - loss: 0.6373
176/176 [==============================] - 1s 5ms/step - loss: 0.6373 - val_loss: 0.4950
Epoch 8/20
1/176 [..............................] - ETA: 0s - loss: 0.5020
13/176 [=>............................] - ETA: 0s - loss: 0.6124
26/176 [===>..........................] - ETA: 0s - loss: 0.5638
39/176 [=====>........................] - ETA: 0s - loss: 0.5533
52/176 [=======>......................] - ETA: 0s - loss: 0.5581
65/176 [==========>...................] - ETA: 0s - loss: 0.5574
78/176 [============>.................] - ETA: 0s - loss: 0.5628
91/176 [==============>...............] - ETA: 0s - loss: 0.5600
105/176 [================>.............] - ETA: 0s - loss: 0.5556
118/176 [===================>..........] - ETA: 0s - loss: 0.5554
131/176 [=====================>........] - ETA: 0s - loss: 0.5513
144/176 [=======================>......] - ETA: 0s - loss: 0.5461
158/176 [=========================>....] - ETA: 0s - loss: 0.5417
172/176 [============================>.] - ETA: 0s - loss: 0.5389
176/176 [==============================] - 1s 4ms/step - loss: 0.5364
176/176 [==============================] - 1s 5ms/step - loss: 0.5364 - val_loss: 0.6750
Epoch 9/20
1/176 [..............................] - ETA: 0s - loss: 0.4314
13/176 [=>............................] - ETA: 0s - loss: 0.5246
26/176 [===>..........................] - ETA: 0s - loss: 0.5084
39/176 [=====>........................] - ETA: 0s - loss: 0.5038
52/176 [=======>......................] - ETA: 0s - loss: 0.4992
65/176 [==========>...................] - ETA: 0s - loss: 0.4905
79/176 [============>.................] - ETA: 0s - loss: 0.4855
93/176 [==============>...............] - ETA: 0s - loss: 0.4803
106/176 [=================>............] - ETA: 0s - loss: 0.4829
119/176 [===================>..........] - ETA: 0s - loss: 0.4763
132/176 [=====================>........] - ETA: 0s - loss: 0.4700
145/176 [=======================>......] - ETA: 0s - loss: 0.4693
158/176 [=========================>....] - ETA: 0s - loss: 0.4667
171/176 [============================>.] - ETA: 0s - loss: 0.4671
176/176 [==============================] - 1s 4ms/step - loss: 0.4675
176/176 [==============================] - 1s 5ms/step - loss: 0.4675 - val_loss: 0.4105
Epoch 10/20
1/176 [..............................] - ETA: 0s - loss: 0.5978
12/176 [=>............................] - ETA: 0s - loss: 0.4532
26/176 [===>..........................] - ETA: 0s - loss: 0.4582
39/176 [=====>........................] - ETA: 0s - loss: 0.4516
52/176 [=======>......................] - ETA: 0s - loss: 0.4483
65/176 [==========>...................] - ETA: 0s - loss: 0.4398
78/176 [============>.................] - ETA: 0s - loss: 0.4412
92/176 [==============>...............] - ETA: 0s - loss: 0.4364
106/176 [=================>............] - ETA: 0s - loss: 0.4341
120/176 [===================>..........] - ETA: 0s - loss: 0.4361
134/176 [=====================>........] - ETA: 0s - loss: 0.4369
147/176 [========================>.....] - ETA: 0s - loss: 0.4334
161/176 [==========================>...] - ETA: 0s - loss: 0.4287
175/176 [============================>.] - ETA: 0s - loss: 0.4274
176/176 [==============================] - 1s 4ms/step - loss: 0.4277
176/176 [==============================] - 1s 5ms/step - loss: 0.4277 - val_loss: 0.5486
Epoch 11/20
1/176 [..............................] - ETA: 2s - loss: 0.4036
12/176 [=>............................] - ETA: 0s - loss: 0.4279
25/176 [===>..........................] - ETA: 0s - loss: 0.3996
38/176 [=====>........................] - ETA: 0s - loss: 0.4026
51/176 [=======>......................] - ETA: 0s - loss: 0.4006
64/176 [=========>....................] - ETA: 0s - loss: 0.4067
77/176 [============>.................] - ETA: 0s - loss: 0.3994
90/176 [==============>...............] - ETA: 0s - loss: 0.3954
104/176 [================>.............] - ETA: 0s - loss: 0.3988
118/176 [===================>..........] - ETA: 0s - loss: 0.3959
131/176 [=====================>........] - ETA: 0s - loss: 0.3964
144/176 [=======================>......] - ETA: 0s - loss: 0.3941
158/176 [=========================>....] - ETA: 0s - loss: 0.3922
172/176 [============================>.] - ETA: 0s - loss: 0.3909
176/176 [==============================] - 1s 4ms/step - loss: 0.3914
176/176 [==============================] - 1s 5ms/step - loss: 0.3914 - val_loss: 0.6018
Epoch 12/20
1/176 [..............................] - ETA: 1s - loss: 0.2622
13/176 [=>............................] - ETA: 0s - loss: 0.3742
26/176 [===>..........................] - ETA: 0s - loss: 0.3630
38/176 [=====>........................] - ETA: 0s - loss: 0.3674
51/176 [=======>......................] - ETA: 0s - loss: 0.3728
64/176 [=========>....................] - ETA: 0s - loss: 0.3690
77/176 [============>.................] - ETA: 0s - loss: 0.3615
91/176 [==============>...............] - ETA: 0s - loss: 0.3632
105/176 [================>.............] - ETA: 0s - loss: 0.3651
118/176 [===================>..........] - ETA: 0s - loss: 0.3594
132/176 [=====================>........] - ETA: 0s - loss: 0.3588
146/176 [=======================>......] - ETA: 0s - loss: 0.3589
154/176 [=========================>....] - ETA: 0s - loss: 0.3589
165/176 [===========================>..] - ETA: 0s - loss: 0.3564
176/176 [==============================] - 1s 4ms/step - loss: 0.3544
176/176 [==============================] - 1s 5ms/step - loss: 0.3544 - val_loss: 0.5790
Epoch 13/20
1/176 [..............................] - ETA: 0s - loss: 0.5461
11/176 [>.............................] - ETA: 0s - loss: 0.3578
20/176 [==>...........................] - ETA: 0s - loss: 0.3339
34/176 [====>.........................] - ETA: 0s - loss: 0.3351
46/176 [======>.......................] - ETA: 0s - loss: 0.3145
59/176 [=========>....................] - ETA: 0s - loss: 0.3111
72/176 [===========>..................] - ETA: 0s - loss: 0.3099
85/176 [=============>................] - ETA: 0s - loss: 0.3140
98/176 [===============>..............] - ETA: 0s - loss: 0.3187
112/176 [==================>...........] - ETA: 0s - loss: 0.3180
126/176 [====================>.........] - ETA: 0s - loss: 0.3186
139/176 [======================>.......] - ETA: 0s - loss: 0.3187
153/176 [=========================>....] - ETA: 0s - loss: 0.3179
167/176 [===========================>..] - ETA: 0s - loss: 0.3183
176/176 [==============================] - 1s 4ms/step - loss: 0.3182
176/176 [==============================] - 1s 5ms/step - loss: 0.3182 - val_loss: 0.7565
Epoch 14/20
1/176 [..............................] - ETA: 0s - loss: 0.3237
13/176 [=>............................] - ETA: 0s - loss: 0.3101
27/176 [===>..........................] - ETA: 0s - loss: 0.3029
41/176 [=====>........................] - ETA: 0s - loss: 0.3137
54/176 [========>.....................] - ETA: 0s - loss: 0.3283
67/176 [==========>...................] - ETA: 0s - loss: 0.3292
81/176 [============>.................] - ETA: 0s - loss: 0.3264
93/176 [==============>...............] - ETA: 0s - loss: 0.3236
107/176 [=================>............] - ETA: 0s - loss: 0.3241
120/176 [===================>..........] - ETA: 0s - loss: 0.3251
134/176 [=====================>........] - ETA: 0s - loss: 0.3256
148/176 [========================>.....] - ETA: 0s - loss: 0.3290
161/176 [==========================>...] - ETA: 0s - loss: 0.3277
174/176 [============================>.] - ETA: 0s - loss: 0.3276
176/176 [==============================] - 1s 4ms/step - loss: 0.3280
176/176 [==============================] - 1s 5ms/step - loss: 0.3280 - val_loss: 0.6888
Epoch 15/20
1/176 [..............................] - ETA: 0s - loss: 0.3653
13/176 [=>............................] - ETA: 0s - loss: 0.3003
26/176 [===>..........................] - ETA: 0s - loss: 0.3035
39/176 [=====>........................] - ETA: 0s - loss: 0.3128
53/176 [========>.....................] - ETA: 0s - loss: 0.3129
66/176 [==========>...................] - ETA: 0s - loss: 0.3105
80/176 [============>.................] - ETA: 0s - loss: 0.3113
93/176 [==============>...............] - ETA: 0s - loss: 0.3145
106/176 [=================>............] - ETA: 0s - loss: 0.3123
120/176 [===================>..........] - ETA: 0s - loss: 0.3174
134/176 [=====================>........] - ETA: 0s - loss: 0.3196
147/176 [========================>.....] - ETA: 0s - loss: 0.3202
160/176 [==========================>...] - ETA: 0s - loss: 0.3213
173/176 [============================>.] - ETA: 0s - loss: 0.3204
176/176 [==============================] - 1s 4ms/step - loss: 0.3211
176/176 [==============================] - 1s 5ms/step - loss: 0.3211 - val_loss: 0.5276
Epoch 16/20
1/176 [..............................] - ETA: 0s - loss: 0.2975
13/176 [=>............................] - ETA: 0s - loss: 0.3418
27/176 [===>..........................] - ETA: 0s - loss: 0.3462
40/176 [=====>........................] - ETA: 0s - loss: 0.3329
53/176 [========>.....................] - ETA: 0s - loss: 0.3473
67/176 [==========>...................] - ETA: 0s - loss: 0.3474
80/176 [============>.................] - ETA: 0s - loss: 0.3464
93/176 [==============>...............] - ETA: 0s - loss: 0.3467
106/176 [=================>............] - ETA: 0s - loss: 0.3464
120/176 [===================>..........] - ETA: 0s - loss: 0.3497
133/176 [=====================>........] - ETA: 0s - loss: 0.3533
147/176 [========================>.....] - ETA: 0s - loss: 0.3524
161/176 [==========================>...] - ETA: 0s - loss: 0.3496
174/176 [============================>.] - ETA: 0s - loss: 0.3506
176/176 [==============================] - 1s 4ms/step - loss: 0.3506
176/176 [==============================] - 1s 5ms/step - loss: 0.3506 - val_loss: 0.5272
Epoch 17/20
1/176 [..............................] - ETA: 0s - loss: 0.3880
14/176 [=>............................] - ETA: 0s - loss: 0.3724
27/176 [===>..........................] - ETA: 0s - loss: 0.3688
40/176 [=====>........................] - ETA: 0s - loss: 0.3645
54/176 [========>.....................] - ETA: 0s - loss: 0.3717
67/176 [==========>...................] - ETA: 0s - loss: 0.3693
75/176 [===========>..................] - ETA: 0s - loss: 0.3671
86/176 [=============>................] - ETA: 0s - loss: 0.3764
99/176 [===============>..............] - ETA: 0s - loss: 0.3748
113/176 [==================>...........] - ETA: 0s - loss: 0.3689
126/176 [====================>.........] - ETA: 0s - loss: 0.3686
139/176 [======================>.......] - ETA: 0s - loss: 0.3714
153/176 [=========================>....] - ETA: 0s - loss: 0.3723
166/176 [===========================>..] - ETA: 0s - loss: 0.3701
176/176 [==============================] - 1s 4ms/step - loss: 0.3701
176/176 [==============================] - 1s 5ms/step - loss: 0.3701 - val_loss: 0.5059
Epoch 18/20
1/176 [..............................] - ETA: 0s - loss: 0.5213
13/176 [=>............................] - ETA: 0s - loss: 0.4270
26/176 [===>..........................] - ETA: 0s - loss: 0.3946
39/176 [=====>........................] - ETA: 0s - loss: 0.4060
53/176 [========>.....................] - ETA: 0s - loss: 0.3995
67/176 [==========>...................] - ETA: 0s - loss: 0.4028
81/176 [============>.................] - ETA: 0s - loss: 0.4019
93/176 [==============>...............] - ETA: 0s - loss: 0.3986
106/176 [=================>............] - ETA: 0s - loss: 0.4056
120/176 [===================>..........] - ETA: 0s - loss: 0.4058
134/176 [=====================>........] - ETA: 0s - loss: 0.4081
148/176 [========================>.....] - ETA: 0s - loss: 0.4117
162/176 [==========================>...] - ETA: 0s - loss: 0.4104
176/176 [==============================] - 1s 4ms/step - loss: 0.4092
176/176 [==============================] - 1s 5ms/step - loss: 0.4092 - val_loss: 0.3343
Epoch 19/20
1/176 [..............................] - ETA: 0s - loss: 0.4185
9/176 [>.............................] - ETA: 1s - loss: 0.4332
18/176 [==>...........................] - ETA: 1s - loss: 0.4030
31/176 [====>.........................] - ETA: 0s - loss: 0.3808
44/176 [======>.......................] - ETA: 0s - loss: 0.3790
58/176 [========>.....................] - ETA: 0s - loss: 0.3800
71/176 [===========>..................] - ETA: 0s - loss: 0.3773
85/176 [=============>................] - ETA: 0s - loss: 0.3760
98/176 [===============>..............] - ETA: 0s - loss: 0.3767
112/176 [==================>...........] - ETA: 0s - loss: 0.3747
126/176 [====================>.........] - ETA: 0s - loss: 0.3746
139/176 [======================>.......] - ETA: 0s - loss: 0.3776
153/176 [=========================>....] - ETA: 0s - loss: 0.3784
167/176 [===========================>..] - ETA: 0s - loss: 0.3819
176/176 [==============================] - 1s 4ms/step - loss: 0.3845
176/176 [==============================] - 1s 5ms/step - loss: 0.3845 - val_loss: 0.3255
Epoch 20/20
1/176 [..............................] - ETA: 0s - loss: 0.4682
13/176 [=>............................] - ETA: 0s - loss: 0.4420
27/176 [===>..........................] - ETA: 0s - loss: 0.4251
40/176 [=====>........................] - ETA: 0s - loss: 0.4128
54/176 [========>.....................] - ETA: 0s - loss: 0.4056
67/176 [==========>...................] - ETA: 0s - loss: 0.4111
81/176 [============>.................] - ETA: 0s - loss: 0.4041
94/176 [===============>..............] - ETA: 0s - loss: 0.4010
108/176 [=================>............] - ETA: 0s - loss: 0.3992
122/176 [===================>..........] - ETA: 0s - loss: 0.4013
135/176 [======================>.......] - ETA: 0s - loss: 0.3937
148/176 [========================>.....] - ETA: 0s - loss: 0.3906
161/176 [==========================>...] - ETA: 0s - loss: 0.3880
174/176 [============================>.] - ETA: 0s - loss: 0.3868
176/176 [==============================] - 1s 4ms/step - loss: 0.3864
176/176 [==============================] - 1s 5ms/step - loss: 0.3864 - val_loss: 0.1994kerasPrediction<-model %>% predict(as.matrix(carbonTestingFinal))2024-05-06 09:54:39.726589: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
1/63 [..............................] - ETA: 4s
35/63 [===============>..............] - ETA: 0s
63/63 [==============================] - 0s 2ms/step
63/63 [==============================] - 0s 2ms/steprmse=function(x,y){
return((mean(x-y)^2)^0.5)
}
rmse(kerasPrediction,carbonTestLabels)[1] 0.2110004MAE(kerasPrediction,carbonTestLabels)[1] 0.2486037rsquared<-sum((kerasPrediction-carbonTestLabels)^2)/sum((carbonTestLabels-mean(carbonTestLabels))^2)
rsquared[1] 0.473026library(tfruns)
runs<-tuning_run(
"carbonEmission.R",
flags=list(
learning_rate=c(0.1,0.5,0.01,0.001),
nodes=c(8,16,32,64,128),
batch_size=c(16,32,64,128),
dropout=c(0.1,0.2,0.3,0.4,0.5),
activation=c("relu")
),sample=0.05
)400 total combinations of flags
(sampled to 20 combinations)yTraining run 1/20 (flags = list(0.5, 128, 128, 0.3, "relu"))
Using run directory runs/2024-05-06T14-54-50Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:54:50.919664: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0102s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0102s). Check your callbacks.
2024-05-06 09:54:53.231240: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 4s - loss: 11.6138 - val_loss: 0.4636 - 4s/epoch - 62ms/step
Epoch 2/20
57/57 - 0s - loss: 2.8228 - val_loss: 0.4192 - 429ms/epoch - 8ms/step
Epoch 3/20
57/57 - 1s - loss: 1.9487 - val_loss: 0.5122 - 965ms/epoch - 17ms/step
Epoch 4/20
57/57 - 1s - loss: 1.4302 - val_loss: 0.3564 - 707ms/epoch - 12ms/step
Epoch 5/20
57/57 - 0s - loss: 1.0963 - val_loss: 0.3395 - 419ms/epoch - 7ms/step
Epoch 6/20
57/57 - 1s - loss: 0.8655 - val_loss: 0.3819 - 699ms/epoch - 12ms/step
Epoch 7/20
57/57 - 1s - loss: 0.6999 - val_loss: 0.2920 - 1s/epoch - 20ms/step
Epoch 8/20
57/57 - 1s - loss: 0.5871 - val_loss: 0.5491 - 704ms/epoch - 12ms/step
Epoch 9/20
57/57 - 0s - loss: 0.5172 - val_loss: 0.4968 - 421ms/epoch - 7ms/step
Epoch 10/20
57/57 - 1s - loss: 0.4489 - val_loss: 0.2608 - 659ms/epoch - 12ms/step
Epoch 11/20
57/57 - 1s - loss: 0.4264 - val_loss: 0.6395 - 698ms/epoch - 12ms/step
Epoch 12/20
57/57 - 1s - loss: 0.3757 - val_loss: 0.5813 - 501ms/epoch - 9ms/step
Epoch 13/20
57/57 - 1s - loss: 0.3551 - val_loss: 0.7431 - 691ms/epoch - 12ms/step
Epoch 14/20
57/57 - 0s - loss: 0.3338 - val_loss: 0.6908 - 445ms/epoch - 8ms/step
Epoch 15/20
57/57 - 1s - loss: 0.3122 - val_loss: 0.6215 - 660ms/epoch - 12ms/step
Epoch 16/20
57/57 - 1s - loss: 0.3136 - val_loss: 0.7349 - 957ms/epoch - 17ms/step
Epoch 17/20
57/57 - 1s - loss: 0.2970 - val_loss: 0.8001 - 908ms/epoch - 16ms/step
Epoch 18/20
57/57 - 0s - loss: 0.2884 - val_loss: 0.9459 - 428ms/epoch - 8ms/step
Epoch 19/20
57/57 - 1s - loss: 0.3036 - val_loss: 0.8340 - 694ms/epoch - 12ms/step
Epoch 20/20
57/57 - 0s - loss: 0.3165 - val_loss: 0.5513 - 455ms/epoch - 8ms/step
Run completed: runs/2024-05-06T14-54-50Z
Training run 2/20 (flags = list(0.5, 8, 128, 0.5, "relu"))
Using run directory runs/2024-05-06T14-55-07Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:55:07.670563: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.3833s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.3833s). Check your callbacks.
2024-05-06 09:55:10.590925: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 4s - loss: 58.2908 - val_loss: 44.7017 - 4s/epoch - 67ms/step
Epoch 2/20
57/57 - 0s - loss: 44.4467 - val_loss: 32.0859 - 434ms/epoch - 8ms/step
Epoch 3/20
57/57 - 1s - loss: 34.2047 - val_loss: 25.5460 - 969ms/epoch - 17ms/step
Epoch 4/20
57/57 - 1s - loss: 29.2576 - val_loss: 24.3943 - 697ms/epoch - 12ms/step
Epoch 5/20
57/57 - 0s - loss: 26.4547 - val_loss: 24.6113 - 410ms/epoch - 7ms/step
Epoch 6/20
57/57 - 1s - loss: 24.1613 - val_loss: 24.0585 - 696ms/epoch - 12ms/step
Epoch 7/20
57/57 - 1s - loss: 21.2079 - val_loss: 25.2184 - 844ms/epoch - 15ms/step
Epoch 8/20
57/57 - 0s - loss: 18.4006 - val_loss: 28.0411 - 422ms/epoch - 7ms/step
Epoch 9/20
57/57 - 1s - loss: 15.9729 - val_loss: 30.0195 - 912ms/epoch - 16ms/step
Epoch 10/20
57/57 - 0s - loss: 14.3609 - val_loss: 31.8455 - 438ms/epoch - 8ms/step
Epoch 11/20
57/57 - 0s - loss: 13.1000 - val_loss: 32.6619 - 414ms/epoch - 7ms/step
Epoch 12/20
57/57 - 1s - loss: 12.2610 - val_loss: 33.3808 - 931ms/epoch - 16ms/step
Epoch 13/20
57/57 - 1s - loss: 11.6125 - val_loss: 34.1868 - 688ms/epoch - 12ms/step
Epoch 14/20
57/57 - 0s - loss: 11.0342 - val_loss: 33.9786 - 424ms/epoch - 7ms/step
Epoch 15/20
57/57 - 1s - loss: 10.6773 - val_loss: 34.4642 - 710ms/epoch - 12ms/step
Epoch 16/20
57/57 - 0s - loss: 10.2095 - val_loss: 34.4586 - 422ms/epoch - 7ms/step
Epoch 17/20
57/57 - 1s - loss: 9.8543 - val_loss: 34.6746 - 699ms/epoch - 12ms/step
Epoch 18/20
57/57 - 1s - loss: 9.6432 - val_loss: 34.5086 - 833ms/epoch - 15ms/step
Epoch 19/20
57/57 - 0s - loss: 9.3112 - val_loss: 34.7509 - 433ms/epoch - 8ms/step
Epoch 20/20
57/57 - 1s - loss: 9.0224 - val_loss: 34.3791 - 698ms/epoch - 12ms/step
Run completed: runs/2024-05-06T14-55-07Z
Training run 3/20 (flags = list(0.5, 16, 16, 0.1, "relu"))
Using run directory runs/2024-05-06T14-55-23Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:55:23.897602: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_end` time: 0.0096s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_end` time: 0.0096s). Check your callbacks.
2024-05-06 09:55:29.146733: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.451/451 - 6s - loss: 7.2410 - val_loss: 1.0509 - 6s/epoch - 14ms/step
Epoch 2/20
451/451 - 3s - loss: 0.9180 - val_loss: 2.2882 - 3s/epoch - 8ms/step
Epoch 3/20
451/451 - 3s - loss: 0.2663 - val_loss: 3.0453 - 3s/epoch - 7ms/step
Epoch 4/20
451/451 - 3s - loss: 0.1253 - val_loss: 3.1514 - 3s/epoch - 7ms/step
Epoch 5/20
451/451 - 3s - loss: 0.0965 - val_loss: 4.3244 - 3s/epoch - 7ms/step
Epoch 6/20
451/451 - 3s - loss: 0.0802 - val_loss: 3.3610 - 3s/epoch - 7ms/step
Epoch 7/20
451/451 - 3s - loss: 0.0667 - val_loss: 3.5523 - 3s/epoch - 7ms/step
Epoch 8/20
451/451 - 3s - loss: 0.0601 - val_loss: 3.6344 - 3s/epoch - 7ms/step
Epoch 9/20
451/451 - 3s - loss: 0.0521 - val_loss: 3.9413 - 3s/epoch - 7ms/step
Epoch 10/20
451/451 - 3s - loss: 0.0468 - val_loss: 3.2592 - 3s/epoch - 7ms/step
Epoch 11/20
451/451 - 3s - loss: 0.0412 - val_loss: 3.7525 - 3s/epoch - 7ms/step
Epoch 12/20
451/451 - 3s - loss: 0.0387 - val_loss: 4.1755 - 3s/epoch - 7ms/step
Epoch 13/20
451/451 - 3s - loss: 0.0367 - val_loss: 3.7769 - 3s/epoch - 7ms/step
Epoch 14/20
451/451 - 3s - loss: 0.0360 - val_loss: 3.3005 - 3s/epoch - 7ms/step
Epoch 15/20
451/451 - 3s - loss: 0.0368 - val_loss: 3.6954 - 3s/epoch - 7ms/step
Epoch 16/20
451/451 - 3s - loss: 0.0402 - val_loss: 3.3595 - 3s/epoch - 6ms/step
Epoch 17/20
451/451 - 3s - loss: 0.0483 - val_loss: 3.6650 - 3s/epoch - 6ms/step
Epoch 18/20
451/451 - 3s - loss: 0.0462 - val_loss: 3.5981 - 3s/epoch - 6ms/step
Epoch 19/20
451/451 - 3s - loss: 0.0549 - val_loss: 3.5653 - 3s/epoch - 6ms/step
Epoch 20/20
451/451 - 3s - loss: 0.0733 - val_loss: 3.1187 - 3s/epoch - 6ms/step
Run completed: runs/2024-05-06T14-55-23Z
Training run 4/20 (flags = list(0.001, 32, 16, 0.3, "relu"))
Using run directory runs/2024-05-06T14-56-27Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:56:30.379454: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_begin` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_begin` time: 0.0334s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_begin` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_begin` time: 0.0334s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0100s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0100s). Check your callbacks.
2024-05-06 09:56:33.239231: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.451/451 - 4s - loss: 8.5746 - val_loss: 1.1113 - 4s/epoch - 8ms/step
Epoch 2/20
451/451 - 3s - loss: 1.4460 - val_loss: 1.0462 - 3s/epoch - 7ms/step
Epoch 3/20
451/451 - 3s - loss: 0.9457 - val_loss: 1.1352 - 3s/epoch - 7ms/step
Epoch 4/20
451/451 - 3s - loss: 0.5666 - val_loss: 1.3286 - 3s/epoch - 6ms/step
Epoch 5/20
451/451 - 3s - loss: 0.3588 - val_loss: 1.2885 - 3s/epoch - 6ms/step
Epoch 6/20
451/451 - 3s - loss: 0.2648 - val_loss: 1.2140 - 3s/epoch - 6ms/step
Epoch 7/20
451/451 - 3s - loss: 0.1982 - val_loss: 1.0452 - 3s/epoch - 6ms/step
Epoch 8/20
451/451 - 3s - loss: 0.1835 - val_loss: 0.7532 - 3s/epoch - 6ms/step
Epoch 9/20
451/451 - 3s - loss: 0.1912 - val_loss: 1.4421 - 3s/epoch - 7ms/step
Epoch 10/20
451/451 - 3s - loss: 0.3299 - val_loss: 1.2238 - 3s/epoch - 6ms/step
Epoch 11/20
451/451 - 3s - loss: 0.5239 - val_loss: 1.2345 - 3s/epoch - 6ms/step
Epoch 12/20
451/451 - 3s - loss: 0.9646 - val_loss: 1.5795 - 3s/epoch - 6ms/step
Epoch 13/20
451/451 - 3s - loss: 1.0706 - val_loss: 1.3688 - 3s/epoch - 6ms/step
Epoch 14/20
451/451 - 3s - loss: 1.5272 - val_loss: 1.5513 - 3s/epoch - 6ms/step
Epoch 15/20
451/451 - 3s - loss: 1.3189 - val_loss: 4.9670 - 3s/epoch - 6ms/step
Epoch 16/20
451/451 - 3s - loss: 0.9311 - val_loss: 6.6251 - 3s/epoch - 6ms/step
Epoch 17/20
451/451 - 3s - loss: 1.0078 - val_loss: 11.9357 - 3s/epoch - 6ms/step
Epoch 18/20
451/451 - 3s - loss: 1.7541 - val_loss: 11.9248 - 3s/epoch - 6ms/step
Epoch 19/20
451/451 - 3s - loss: 1.6260 - val_loss: 12.3699 - 3s/epoch - 6ms/step
Epoch 20/20
451/451 - 3s - loss: 1.9291 - val_loss: 20.7996 - 3s/epoch - 6ms/step
Run completed: runs/2024-05-06T14-56-27Z
Training run 5/20 (flags = list(0.1, 32, 128, 0.3, "relu"))
Using run directory runs/2024-05-06T14-57-29Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:57:29.769069: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.1481s vs `on_train_batch_end` time: 0.1940s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.1481s vs `on_train_batch_end` time: 0.1940s). Check your callbacks.
2024-05-06 09:57:32.796985: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 4s - loss: 27.3645 - val_loss: 3.1691 - 4s/epoch - 67ms/step
Epoch 2/20
57/57 - 0s - loss: 9.7090 - val_loss: 2.9526 - 431ms/epoch - 8ms/step
Epoch 3/20
57/57 - 1s - loss: 7.3945 - val_loss: 3.3290 - 862ms/epoch - 15ms/step
Epoch 4/20
57/57 - 1s - loss: 6.3382 - val_loss: 3.0386 - 671ms/epoch - 12ms/step
Epoch 5/20
57/57 - 0s - loss: 5.4127 - val_loss: 2.7711 - 423ms/epoch - 7ms/step
Epoch 6/20
57/57 - 1s - loss: 4.6305 - val_loss: 2.7949 - 657ms/epoch - 12ms/step
Epoch 7/20
57/57 - 0s - loss: 3.7744 - val_loss: 2.5571 - 421ms/epoch - 7ms/step
Epoch 8/20
57/57 - 1s - loss: 3.1616 - val_loss: 2.8495 - 826ms/epoch - 14ms/step
Epoch 9/20
57/57 - 1s - loss: 2.7162 - val_loss: 3.3917 - 764ms/epoch - 13ms/step
Epoch 10/20
57/57 - 0s - loss: 2.4613 - val_loss: 3.4641 - 421ms/epoch - 7ms/step
Epoch 11/20
57/57 - 1s - loss: 2.1948 - val_loss: 4.1984 - 662ms/epoch - 12ms/step
Epoch 12/20
57/57 - 0s - loss: 1.8854 - val_loss: 3.9335 - 417ms/epoch - 7ms/step
Epoch 13/20
57/57 - 1s - loss: 1.6623 - val_loss: 4.9280 - 662ms/epoch - 12ms/step
Epoch 14/20
57/57 - 0s - loss: 1.5136 - val_loss: 5.0246 - 421ms/epoch - 7ms/step
Epoch 15/20
57/57 - 1s - loss: 1.4015 - val_loss: 5.3457 - 621ms/epoch - 11ms/step
Epoch 16/20
57/57 - 0s - loss: 1.2787 - val_loss: 5.9332 - 420ms/epoch - 7ms/step
Epoch 17/20
57/57 - 0s - loss: 1.1320 - val_loss: 5.2903 - 418ms/epoch - 7ms/step
Epoch 18/20
57/57 - 1s - loss: 1.0409 - val_loss: 5.9916 - 972ms/epoch - 17ms/step
Epoch 19/20
57/57 - 0s - loss: 0.9905 - val_loss: 5.5906 - 451ms/epoch - 8ms/step
Epoch 20/20
57/57 - 1s - loss: 0.8973 - val_loss: 5.6969 - 634ms/epoch - 11ms/step
Run completed: runs/2024-05-06T14-57-29Z
Training run 6/20 (flags = list(0.1, 8, 128, 0.5, "relu"))
Using run directory runs/2024-05-06T14-57-44Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:57:46.559609: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0043s vs `on_train_batch_end` time: 0.0100s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0043s vs `on_train_batch_end` time: 0.0100s). Check your callbacks.
2024-05-06 09:57:47.701597: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 4s - loss: 54.2360 - val_loss: 39.9696 - 4s/epoch - 63ms/step
Epoch 2/20
57/57 - 0s - loss: 42.4228 - val_loss: 29.1418 - 440ms/epoch - 8ms/step
Epoch 3/20
57/57 - 1s - loss: 35.7827 - val_loss: 24.1368 - 823ms/epoch - 14ms/step
Epoch 4/20
57/57 - 0s - loss: 31.3284 - val_loss: 21.2503 - 439ms/epoch - 8ms/step
Epoch 5/20
57/57 - 1s - loss: 28.3253 - val_loss: 20.8096 - 613ms/epoch - 11ms/step
Epoch 6/20
57/57 - 1s - loss: 25.3994 - val_loss: 20.4552 - 642ms/epoch - 11ms/step
Epoch 7/20
57/57 - 0s - loss: 23.2297 - val_loss: 21.8989 - 422ms/epoch - 7ms/step
Epoch 8/20
57/57 - 1s - loss: 21.2513 - val_loss: 22.9363 - 1s/epoch - 18ms/step
Epoch 9/20
57/57 - 0s - loss: 20.3645 - val_loss: 24.5281 - 415ms/epoch - 7ms/step
Epoch 10/20
57/57 - 0s - loss: 19.6282 - val_loss: 24.2801 - 413ms/epoch - 7ms/step
Epoch 11/20
57/57 - 1s - loss: 18.5731 - val_loss: 24.0698 - 643ms/epoch - 11ms/step
Epoch 12/20
57/57 - 0s - loss: 17.1591 - val_loss: 23.1534 - 417ms/epoch - 7ms/step
Epoch 13/20
57/57 - 1s - loss: 16.1036 - val_loss: 22.8076 - 839ms/epoch - 15ms/step
Epoch 14/20
57/57 - 0s - loss: 15.0624 - val_loss: 22.3049 - 416ms/epoch - 7ms/step
Epoch 15/20
57/57 - 1s - loss: 14.0257 - val_loss: 21.9625 - 617ms/epoch - 11ms/step
Epoch 16/20
57/57 - 0s - loss: 13.3588 - val_loss: 21.1181 - 409ms/epoch - 7ms/step
Epoch 17/20
57/57 - 1s - loss: 12.5951 - val_loss: 20.8273 - 624ms/epoch - 11ms/step
Epoch 18/20
57/57 - 1s - loss: 12.1326 - val_loss: 19.8063 - 835ms/epoch - 15ms/step
Epoch 19/20
57/57 - 0s - loss: 11.6216 - val_loss: 19.3653 - 455ms/epoch - 8ms/step
Epoch 20/20
57/57 - 1s - loss: 11.2639 - val_loss: 19.1199 - 662ms/epoch - 12ms/step
Run completed: runs/2024-05-06T14-57-44Z
Training run 7/20 (flags = list(0.001, 8, 64, 0.2, "relu"))
Using run directory runs/2024-05-06T14-57-59Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:58:00.117053: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0051s vs `on_train_batch_end` time: 0.0661s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0051s vs `on_train_batch_end` time: 0.0661s). Check your callbacks.
2024-05-06 09:58:03.043125: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.113/113 - 4s - loss: 27.8824 - val_loss: 10.0069 - 4s/epoch - 33ms/step
Epoch 2/20
113/113 - 1s - loss: 14.4011 - val_loss: 7.1101 - 1s/epoch - 11ms/step
Epoch 3/20
113/113 - 1s - loss: 10.9057 - val_loss: 6.6431 - 933ms/epoch - 8ms/step
Epoch 4/20
113/113 - 1s - loss: 7.8081 - val_loss: 7.3311 - 944ms/epoch - 8ms/step
Epoch 5/20
113/113 - 1s - loss: 4.8643 - val_loss: 8.3484 - 1s/epoch - 10ms/step
Epoch 6/20
113/113 - 1s - loss: 3.5000 - val_loss: 8.6690 - 1s/epoch - 10ms/step
Epoch 7/20
113/113 - 1s - loss: 2.9008 - val_loss: 8.8686 - 706ms/epoch - 6ms/step
Epoch 8/20
113/113 - 1s - loss: 2.5954 - val_loss: 8.8508 - 930ms/epoch - 8ms/step
Epoch 9/20
113/113 - 1s - loss: 2.3520 - val_loss: 8.8799 - 922ms/epoch - 8ms/step
Epoch 10/20
113/113 - 1s - loss: 2.2080 - val_loss: 9.0141 - 927ms/epoch - 8ms/step
Epoch 11/20
113/113 - 1s - loss: 2.0546 - val_loss: 8.7873 - 1s/epoch - 10ms/step
Epoch 12/20
113/113 - 1s - loss: 1.8869 - val_loss: 9.1333 - 1s/epoch - 10ms/step
Epoch 13/20
113/113 - 1s - loss: 1.7640 - val_loss: 9.3086 - 950ms/epoch - 8ms/step
Epoch 14/20
113/113 - 1s - loss: 1.6231 - val_loss: 9.0543 - 935ms/epoch - 8ms/step
Epoch 15/20
113/113 - 1s - loss: 1.5163 - val_loss: 9.3269 - 939ms/epoch - 8ms/step
Epoch 16/20
113/113 - 1s - loss: 1.4586 - val_loss: 9.2640 - 933ms/epoch - 8ms/step
Epoch 17/20
113/113 - 1s - loss: 1.3899 - val_loss: 9.9020 - 936ms/epoch - 8ms/step
Epoch 18/20
113/113 - 1s - loss: 1.2809 - val_loss: 9.8456 - 891ms/epoch - 8ms/step
Epoch 19/20
113/113 - 1s - loss: 1.2347 - val_loss: 9.3953 - 703ms/epoch - 6ms/step
Epoch 20/20
113/113 - 1s - loss: 1.2234 - val_loss: 9.6033 - 888ms/epoch - 8ms/step
Run completed: runs/2024-05-06T14-57-59Z
Training run 8/20 (flags = list(0.5, 32, 32, 0.3, "relu"))
Using run directory runs/2024-05-06T14-58-22Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:58:22.755810: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0041s vs `on_train_batch_end` time: 0.0099s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0041s vs `on_train_batch_end` time: 0.0099s). Check your callbacks.
2024-05-06 09:58:26.008537: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 4s - loss: 14.1427 - val_loss: 1.4044 - 4s/epoch - 18ms/step
Epoch 2/20
226/226 - 2s - loss: 3.3903 - val_loss: 0.7778 - 2s/epoch - 9ms/step
Epoch 3/20
226/226 - 2s - loss: 1.4087 - val_loss: 1.1622 - 2s/epoch - 9ms/step
Epoch 4/20
226/226 - 1s - loss: 0.8784 - val_loss: 1.3062 - 1s/epoch - 7ms/step
Epoch 5/20
226/226 - 1s - loss: 0.6745 - val_loss: 1.4484 - 1s/epoch - 7ms/step
Epoch 6/20
226/226 - 2s - loss: 0.5957 - val_loss: 1.7683 - 2s/epoch - 7ms/step
Epoch 7/20
226/226 - 1s - loss: 0.5664 - val_loss: 2.2544 - 1s/epoch - 7ms/step
Epoch 8/20
226/226 - 2s - loss: 0.5056 - val_loss: 3.0286 - 2s/epoch - 8ms/step
Epoch 9/20
226/226 - 2s - loss: 0.4943 - val_loss: 3.6171 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 2s - loss: 0.4270 - val_loss: 4.5025 - 2s/epoch - 8ms/step
Epoch 11/20
226/226 - 2s - loss: 0.3557 - val_loss: 6.1073 - 2s/epoch - 8ms/step
Epoch 12/20
226/226 - 2s - loss: 0.2998 - val_loss: 6.1265 - 2s/epoch - 7ms/step
Epoch 13/20
226/226 - 2s - loss: 0.2564 - val_loss: 5.8923 - 2s/epoch - 8ms/step
Epoch 14/20
226/226 - 2s - loss: 0.2483 - val_loss: 7.0588 - 2s/epoch - 8ms/step
Epoch 15/20
226/226 - 1s - loss: 0.2262 - val_loss: 7.1452 - 1s/epoch - 7ms/step
Epoch 16/20
226/226 - 2s - loss: 0.2271 - val_loss: 6.2520 - 2s/epoch - 7ms/step
Epoch 17/20
226/226 - 2s - loss: 0.2006 - val_loss: 6.2938 - 2s/epoch - 7ms/step
Epoch 18/20
226/226 - 2s - loss: 0.1824 - val_loss: 6.5048 - 2s/epoch - 7ms/step
Epoch 19/20
226/226 - 1s - loss: 0.1703 - val_loss: 5.4294 - 1s/epoch - 7ms/step
Epoch 20/20
226/226 - 2s - loss: 0.2027 - val_loss: 4.8250 - 2s/epoch - 8ms/step
Run completed: runs/2024-05-06T14-58-22Z
Training run 9/20 (flags = list(0.5, 64, 64, 0.5, "relu"))
Using run directory runs/2024-05-06T14-58-58Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:58:58.587922: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0089s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0047s vs `on_train_batch_end` time: 0.0089s). Check your callbacks.
2024-05-06 09:59:01.622113: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.113/113 - 4s - loss: 16.3675 - val_loss: 4.4610 - 4s/epoch - 37ms/step
Epoch 2/20
113/113 - 1s - loss: 4.3639 - val_loss: 2.9391 - 1s/epoch - 9ms/step
Epoch 3/20
113/113 - 1s - loss: 2.3112 - val_loss: 2.8172 - 844ms/epoch - 7ms/step
Epoch 4/20
113/113 - 1s - loss: 1.7284 - val_loss: 3.3120 - 714ms/epoch - 6ms/step
Epoch 5/20
113/113 - 1s - loss: 1.3500 - val_loss: 2.7280 - 1s/epoch - 12ms/step
Epoch 6/20
113/113 - 1s - loss: 1.2119 - val_loss: 3.1502 - 1s/epoch - 10ms/step
Epoch 7/20
113/113 - 1s - loss: 1.0716 - val_loss: 3.1872 - 1s/epoch - 10ms/step
Epoch 8/20
113/113 - 1s - loss: 0.9595 - val_loss: 3.0978 - 931ms/epoch - 8ms/step
Epoch 9/20
113/113 - 1s - loss: 0.8513 - val_loss: 2.8730 - 937ms/epoch - 8ms/step
Epoch 10/20
113/113 - 1s - loss: 0.8251 - val_loss: 2.8990 - 922ms/epoch - 8ms/step
Epoch 11/20
113/113 - 1s - loss: 0.7660 - val_loss: 2.9483 - 1s/epoch - 12ms/step
Epoch 12/20
113/113 - 1s - loss: 0.6602 - val_loss: 2.7560 - 940ms/epoch - 8ms/step
Epoch 13/20
113/113 - 1s - loss: 0.5844 - val_loss: 2.8620 - 1s/epoch - 10ms/step
Epoch 14/20
113/113 - 1s - loss: 0.5224 - val_loss: 3.1427 - 1s/epoch - 10ms/step
Epoch 15/20
113/113 - 1s - loss: 0.5022 - val_loss: 3.1201 - 696ms/epoch - 6ms/step
Epoch 16/20
113/113 - 1s - loss: 0.4721 - val_loss: 3.0309 - 946ms/epoch - 8ms/step
Epoch 17/20
113/113 - 1s - loss: 0.4312 - val_loss: 3.3700 - 944ms/epoch - 8ms/step
Epoch 18/20
113/113 - 1s - loss: 0.3745 - val_loss: 3.1685 - 930ms/epoch - 8ms/step
Epoch 19/20
113/113 - 1s - loss: 0.3299 - val_loss: 3.6113 - 1s/epoch - 10ms/step
Epoch 20/20
113/113 - 1s - loss: 0.3444 - val_loss: 3.0329 - 1s/epoch - 10ms/step
Run completed: runs/2024-05-06T14-58-58Z
Training run 10/20 (flags = list(0.001, 32, 16, 0.1, "relu"))
Using run directory runs/2024-05-06T14-59-22Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 09:59:22.860434: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0048s vs `on_train_batch_end` time: 0.0682s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0048s vs `on_train_batch_end` time: 0.0682s). Check your callbacks.
2024-05-06 09:59:27.537747: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.451/451 - 6s - loss: 5.4132 - val_loss: 2.2404 - 6s/epoch - 13ms/step
Epoch 2/20
451/451 - 3s - loss: 0.2787 - val_loss: 2.8976 - 3s/epoch - 7ms/step
Epoch 3/20
451/451 - 3s - loss: 0.1131 - val_loss: 3.4428 - 3s/epoch - 7ms/step
Epoch 4/20
451/451 - 3s - loss: 0.0763 - val_loss: 3.0443 - 3s/epoch - 6ms/step
Epoch 5/20
451/451 - 3s - loss: 0.0570 - val_loss: 3.2611 - 3s/epoch - 6ms/step
Epoch 6/20
451/451 - 3s - loss: 0.0493 - val_loss: 3.5065 - 3s/epoch - 6ms/step
Epoch 7/20
451/451 - 3s - loss: 0.0466 - val_loss: 3.2422 - 3s/epoch - 6ms/step
Epoch 8/20
451/451 - 3s - loss: 0.0528 - val_loss: 3.9538 - 3s/epoch - 6ms/step
Epoch 9/20
451/451 - 3s - loss: 0.1137 - val_loss: 4.0139 - 3s/epoch - 7ms/step
Epoch 10/20
451/451 - 3s - loss: 0.3051 - val_loss: 7.7234 - 3s/epoch - 6ms/step
Epoch 11/20
451/451 - 3s - loss: 0.6024 - val_loss: 4.9395 - 3s/epoch - 6ms/step
Epoch 12/20
451/451 - 3s - loss: 0.4960 - val_loss: 5.7518 - 3s/epoch - 6ms/step
Epoch 13/20
451/451 - 3s - loss: 0.6740 - val_loss: 6.0071 - 3s/epoch - 6ms/step
Epoch 14/20
451/451 - 3s - loss: 0.8054 - val_loss: 5.9459 - 3s/epoch - 6ms/step
Epoch 15/20
451/451 - 3s - loss: 0.8653 - val_loss: 7.6197 - 3s/epoch - 6ms/step
Epoch 16/20
451/451 - 3s - loss: 0.9881 - val_loss: 7.2715 - 3s/epoch - 7ms/step
Epoch 17/20
451/451 - 3s - loss: 1.7702 - val_loss: 4.9933 - 3s/epoch - 7ms/step
Epoch 18/20
451/451 - 3s - loss: 0.8956 - val_loss: 7.7228 - 3s/epoch - 7ms/step
Epoch 19/20
451/451 - 3s - loss: 1.1377 - val_loss: 6.0371 - 3s/epoch - 7ms/step
Epoch 20/20
451/451 - 3s - loss: 0.9618 - val_loss: 4.6644 - 3s/epoch - 7ms/step
Run completed: runs/2024-05-06T14-59-22Z
Training run 11/20 (flags = list(0.001, 128, 128, 0.1, "relu"))
Using run directory runs/2024-05-06T15-00-23Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:00:24.091839: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_begin` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_begin` time: 0.0261s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_begin` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_begin` time: 0.0261s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_end` time: 0.2079s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0053s vs `on_train_batch_end` time: 0.2079s). Check your callbacks.
2024-05-06 10:00:26.701102: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 4s - loss: 6.7932 - val_loss: 0.3520 - 4s/epoch - 62ms/step
Epoch 2/20
57/57 - 0s - loss: 1.1441 - val_loss: 0.1199 - 433ms/epoch - 8ms/step
Epoch 3/20
57/57 - 1s - loss: 0.9673 - val_loss: 0.0774 - 659ms/epoch - 12ms/step
Epoch 4/20
57/57 - 0s - loss: 0.8415 - val_loss: 0.0567 - 419ms/epoch - 7ms/step
Epoch 5/20
57/57 - 1s - loss: 0.7376 - val_loss: 0.0749 - 646ms/epoch - 11ms/step
Epoch 6/20
57/57 - 0s - loss: 0.6615 - val_loss: 0.0482 - 421ms/epoch - 7ms/step
Epoch 7/20
57/57 - 1s - loss: 0.5997 - val_loss: 0.0830 - 845ms/epoch - 15ms/step
Epoch 8/20
57/57 - 1s - loss: 0.5506 - val_loss: 0.0730 - 653ms/epoch - 11ms/step
Epoch 9/20
57/57 - 0s - loss: 0.4975 - val_loss: 0.1078 - 429ms/epoch - 8ms/step
Epoch 10/20
57/57 - 1s - loss: 0.4809 - val_loss: 0.0800 - 852ms/epoch - 15ms/step
Epoch 11/20
57/57 - 0s - loss: 0.4315 - val_loss: 0.0916 - 435ms/epoch - 8ms/step
Epoch 12/20
57/57 - 1s - loss: 0.4245 - val_loss: 0.0799 - 611ms/epoch - 11ms/step
Epoch 13/20
57/57 - 1s - loss: 0.4424 - val_loss: 0.0902 - 806ms/epoch - 14ms/step
Epoch 14/20
57/57 - 0s - loss: 0.4757 - val_loss: 0.1759 - 422ms/epoch - 7ms/step
Epoch 15/20
57/57 - 1s - loss: 0.5080 - val_loss: 0.0634 - 844ms/epoch - 15ms/step
Epoch 16/20
57/57 - 0s - loss: 0.6966 - val_loss: 0.1159 - 423ms/epoch - 7ms/step
Epoch 17/20
57/57 - 0s - loss: 0.8899 - val_loss: 0.1272 - 420ms/epoch - 7ms/step
Epoch 18/20
57/57 - 1s - loss: 1.2849 - val_loss: 0.7818 - 836ms/epoch - 15ms/step
Epoch 19/20
57/57 - 0s - loss: 1.5460 - val_loss: 0.1634 - 475ms/epoch - 8ms/step
Epoch 20/20
57/57 - 1s - loss: 1.6178 - val_loss: 0.8310 - 595ms/epoch - 10ms/step
Run completed: runs/2024-05-06T15-00-23Z
Training run 12/20 (flags = list(0.1, 16, 32, 0.5, "relu"))
Using run directory runs/2024-05-06T15-00-38Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:00:39.227202: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0050s vs `on_train_batch_end` time: 0.0983s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0050s vs `on_train_batch_end` time: 0.0983s). Check your callbacks.
2024-05-06 10:00:42.673791: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 5s - loss: 42.3642 - val_loss: 17.8899 - 5s/epoch - 21ms/step
Epoch 2/20
226/226 - 1s - loss: 16.8195 - val_loss: 16.6061 - 1s/epoch - 7ms/step
Epoch 3/20
226/226 - 2s - loss: 9.7924 - val_loss: 15.2268 - 2s/epoch - 8ms/step
Epoch 4/20
226/226 - 2s - loss: 6.3296 - val_loss: 15.9971 - 2s/epoch - 8ms/step
Epoch 5/20
226/226 - 2s - loss: 4.5965 - val_loss: 15.7332 - 2s/epoch - 8ms/step
Epoch 6/20
226/226 - 2s - loss: 3.5200 - val_loss: 16.1088 - 2s/epoch - 8ms/step
Epoch 7/20
226/226 - 2s - loss: 3.1163 - val_loss: 15.4500 - 2s/epoch - 7ms/step
Epoch 8/20
226/226 - 2s - loss: 2.9552 - val_loss: 16.8626 - 2s/epoch - 8ms/step
Epoch 9/20
226/226 - 2s - loss: 2.9407 - val_loss: 18.6550 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 1s - loss: 2.9194 - val_loss: 20.8167 - 1s/epoch - 7ms/step
Epoch 11/20
226/226 - 2s - loss: 2.6893 - val_loss: 21.3925 - 2s/epoch - 7ms/step
Epoch 12/20
226/226 - 1s - loss: 2.3486 - val_loss: 18.9689 - 1s/epoch - 7ms/step
Epoch 13/20
226/226 - 2s - loss: 2.0517 - val_loss: 19.2253 - 2s/epoch - 7ms/step
Epoch 14/20
226/226 - 2s - loss: 1.6478 - val_loss: 20.3939 - 2s/epoch - 8ms/step
Epoch 15/20
226/226 - 2s - loss: 1.4001 - val_loss: 15.6542 - 2s/epoch - 7ms/step
Epoch 16/20
226/226 - 2s - loss: 1.2497 - val_loss: 16.5755 - 2s/epoch - 7ms/step
Epoch 17/20
226/226 - 2s - loss: 0.8850 - val_loss: 16.2725 - 2s/epoch - 9ms/step
Epoch 18/20
226/226 - 2s - loss: 0.7459 - val_loss: 15.4820 - 2s/epoch - 8ms/step
Epoch 19/20
226/226 - 2s - loss: 0.7532 - val_loss: 15.7535 - 2s/epoch - 8ms/step
Epoch 20/20
226/226 - 1s - loss: 0.7799 - val_loss: 15.0892 - 1s/epoch - 7ms/step
Run completed: runs/2024-05-06T15-00-38Z
Training run 13/20 (flags = list(0.001, 16, 128, 0.1, "relu"))
Using run directory runs/2024-05-06T15-01-16Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:01:17.193063: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0051s vs `on_train_batch_end` time: 0.0104s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0051s vs `on_train_batch_end` time: 0.0104s). Check your callbacks.
2024-05-06 10:01:19.315139: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 3s - loss: 32.8509 - val_loss: 6.3026 - 3s/epoch - 51ms/step
Epoch 2/20
57/57 - 1s - loss: 7.4565 - val_loss: 1.1933 - 1s/epoch - 20ms/step
Epoch 3/20
57/57 - 1s - loss: 5.2598 - val_loss: 0.8431 - 832ms/epoch - 15ms/step
Epoch 4/20
57/57 - 0s - loss: 4.5574 - val_loss: 0.8813 - 415ms/epoch - 7ms/step
Epoch 5/20
57/57 - 0s - loss: 4.1343 - val_loss: 0.7815 - 414ms/epoch - 7ms/step
Epoch 6/20
57/57 - 1s - loss: 3.7398 - val_loss: 0.7278 - 778ms/epoch - 14ms/step
Epoch 7/20
57/57 - 0s - loss: 3.3985 - val_loss: 0.7768 - 432ms/epoch - 8ms/step
Epoch 8/20
57/57 - 1s - loss: 3.1007 - val_loss: 0.7981 - 659ms/epoch - 12ms/step
Epoch 9/20
57/57 - 0s - loss: 2.7546 - val_loss: 0.9694 - 417ms/epoch - 7ms/step
Epoch 10/20
57/57 - 1s - loss: 2.5242 - val_loss: 1.1810 - 855ms/epoch - 15ms/step
Epoch 11/20
57/57 - 0s - loss: 2.3096 - val_loss: 1.2593 - 422ms/epoch - 7ms/step
Epoch 12/20
57/57 - 0s - loss: 2.0929 - val_loss: 1.4915 - 423ms/epoch - 7ms/step
Epoch 13/20
57/57 - 0s - loss: 1.7844 - val_loss: 1.6409 - 416ms/epoch - 7ms/step
Epoch 14/20
57/57 - 1s - loss: 1.5349 - val_loss: 1.7480 - 836ms/epoch - 15ms/step
Epoch 15/20
57/57 - 0s - loss: 1.2141 - val_loss: 2.3716 - 424ms/epoch - 7ms/step
Epoch 16/20
57/57 - 0s - loss: 1.0039 - val_loss: 2.7482 - 422ms/epoch - 7ms/step
Epoch 17/20
57/57 - 0s - loss: 0.7942 - val_loss: 2.4333 - 443ms/epoch - 8ms/step
Epoch 18/20
57/57 - 1s - loss: 0.6808 - val_loss: 3.0976 - 623ms/epoch - 11ms/step
Epoch 19/20
57/57 - 1s - loss: 0.5678 - val_loss: 3.3911 - 832ms/epoch - 15ms/step
Epoch 20/20
57/57 - 0s - loss: 0.5200 - val_loss: 3.8461 - 444ms/epoch - 8ms/step
Run completed: runs/2024-05-06T15-01-16Z
Training run 14/20 (flags = list(0.5, 128, 32, 0.5, "relu"))
Using run directory runs/2024-05-06T15-01-31Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:01:31.902364: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0044s vs `on_train_batch_end` time: 0.0093s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0044s vs `on_train_batch_end` time: 0.0093s). Check your callbacks.
2024-05-06 10:01:35.249219: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 4s - loss: 6.7435 - val_loss: 2.2936 - 4s/epoch - 18ms/step
Epoch 2/20
226/226 - 2s - loss: 0.8185 - val_loss: 2.3855 - 2s/epoch - 8ms/step
Epoch 3/20
226/226 - 2s - loss: 0.4637 - val_loss: 2.3568 - 2s/epoch - 8ms/step
Epoch 4/20
226/226 - 2s - loss: 0.3194 - val_loss: 2.4118 - 2s/epoch - 9ms/step
Epoch 5/20
226/226 - 2s - loss: 0.2394 - val_loss: 1.9110 - 2s/epoch - 7ms/step
Epoch 6/20
226/226 - 2s - loss: 0.1977 - val_loss: 2.2398 - 2s/epoch - 9ms/step
Epoch 7/20
226/226 - 2s - loss: 0.1703 - val_loss: 1.7173 - 2s/epoch - 9ms/step
Epoch 8/20
226/226 - 2s - loss: 0.1412 - val_loss: 1.7953 - 2s/epoch - 9ms/step
Epoch 9/20
226/226 - 2s - loss: 0.1579 - val_loss: 1.9782 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 2s - loss: 0.1898 - val_loss: 2.3954 - 2s/epoch - 7ms/step
Epoch 11/20
226/226 - 2s - loss: 0.3629 - val_loss: 2.2133 - 2s/epoch - 9ms/step
Epoch 12/20
226/226 - 2s - loss: 0.2497 - val_loss: 2.1539 - 2s/epoch - 9ms/step
Epoch 13/20
226/226 - 2s - loss: 0.5479 - val_loss: 1.7509 - 2s/epoch - 9ms/step
Epoch 14/20
226/226 - 2s - loss: 0.7239 - val_loss: 1.6447 - 2s/epoch - 9ms/step
Epoch 15/20
226/226 - 2s - loss: 0.7419 - val_loss: 2.2293 - 2s/epoch - 7ms/step
Epoch 16/20
226/226 - 2s - loss: 0.6668 - val_loss: 5.7126 - 2s/epoch - 7ms/step
Epoch 17/20
226/226 - 2s - loss: 1.0296 - val_loss: 4.6779 - 2s/epoch - 9ms/step
Epoch 18/20
226/226 - 2s - loss: 3.1392 - val_loss: 28.5436 - 2s/epoch - 9ms/step
Epoch 19/20
226/226 - 2s - loss: 1.2982 - val_loss: 35.5697 - 2s/epoch - 9ms/step
Epoch 20/20
226/226 - 2s - loss: 1.7788 - val_loss: 31.6501 - 2s/epoch - 8ms/step
Run completed: runs/2024-05-06T15-01-31Z
Training run 15/20 (flags = list(0.5, 16, 32, 0.2, "relu"))
Using run directory runs/2024-05-06T15-02-11Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:02:11.399961: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0048s vs `on_train_batch_end` time: 0.0979s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0048s vs `on_train_batch_end` time: 0.0979s). Check your callbacks.
2024-05-06 10:02:14.933991: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 5s - loss: 15.1082 - val_loss: 2.4513 - 5s/epoch - 21ms/step
Epoch 2/20
226/226 - 2s - loss: 4.0246 - val_loss: 2.3613 - 2s/epoch - 7ms/step
Epoch 3/20
226/226 - 1s - loss: 2.3179 - val_loss: 2.7398 - 1s/epoch - 7ms/step
Epoch 4/20
226/226 - 2s - loss: 1.4267 - val_loss: 3.9060 - 2s/epoch - 9ms/step
Epoch 5/20
226/226 - 2s - loss: 0.9689 - val_loss: 4.5371 - 2s/epoch - 7ms/step
Epoch 6/20
226/226 - 2s - loss: 0.6731 - val_loss: 4.8713 - 2s/epoch - 7ms/step
Epoch 7/20
226/226 - 2s - loss: 0.4922 - val_loss: 5.6833 - 2s/epoch - 8ms/step
Epoch 8/20
226/226 - 2s - loss: 0.4046 - val_loss: 6.0810 - 2s/epoch - 8ms/step
Epoch 9/20
226/226 - 2s - loss: 0.3344 - val_loss: 5.9184 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 2s - loss: 0.2714 - val_loss: 6.4398 - 2s/epoch - 8ms/step
Epoch 11/20
226/226 - 1s - loss: 0.2485 - val_loss: 6.7110 - 1s/epoch - 7ms/step
Epoch 12/20
226/226 - 2s - loss: 0.2161 - val_loss: 6.0764 - 2s/epoch - 7ms/step
Epoch 13/20
226/226 - 2s - loss: 0.1849 - val_loss: 7.1760 - 2s/epoch - 8ms/step
Epoch 14/20
226/226 - 1s - loss: 0.1611 - val_loss: 6.3526 - 1s/epoch - 6ms/step
Epoch 15/20
226/226 - 1s - loss: 0.1461 - val_loss: 7.2512 - 1s/epoch - 6ms/step
Epoch 16/20
226/226 - 2s - loss: 0.1321 - val_loss: 7.5646 - 2s/epoch - 7ms/step
Epoch 17/20
226/226 - 2s - loss: 0.1162 - val_loss: 6.9163 - 2s/epoch - 7ms/step
Epoch 18/20
226/226 - 2s - loss: 0.1094 - val_loss: 6.8628 - 2s/epoch - 7ms/step
Epoch 19/20
226/226 - 1s - loss: 0.0999 - val_loss: 7.8049 - 1s/epoch - 6ms/step
Epoch 20/20
226/226 - 2s - loss: 0.0961 - val_loss: 6.8945 - 2s/epoch - 7ms/step
Run completed: runs/2024-05-06T15-02-11Z
Training run 16/20 (flags = list(0.001, 8, 32, 0.3, "relu"))
Using run directory runs/2024-05-06T15-02-47Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:02:47.420338: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0060s vs `on_train_batch_end` time: 0.2036s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0060s vs `on_train_batch_end` time: 0.2036s). Check your callbacks.
2024-05-06 10:02:51.137654: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 5s - loss: 29.1013 - val_loss: 11.0376 - 5s/epoch - 21ms/step
Epoch 2/20
226/226 - 2s - loss: 10.5182 - val_loss: 13.0675 - 2s/epoch - 7ms/step
Epoch 3/20
226/226 - 2s - loss: 5.5168 - val_loss: 18.3696 - 2s/epoch - 8ms/step
Epoch 4/20
226/226 - 2s - loss: 4.0513 - val_loss: 18.9510 - 2s/epoch - 9ms/step
Epoch 5/20
226/226 - 2s - loss: 3.0583 - val_loss: 20.2050 - 2s/epoch - 8ms/step
Epoch 6/20
226/226 - 2s - loss: 2.1942 - val_loss: 21.3651 - 2s/epoch - 7ms/step
Epoch 7/20
226/226 - 2s - loss: 1.6739 - val_loss: 22.8143 - 2s/epoch - 8ms/step
Epoch 8/20
226/226 - 2s - loss: 1.4806 - val_loss: 22.5425 - 2s/epoch - 8ms/step
Epoch 9/20
226/226 - 2s - loss: 1.2700 - val_loss: 23.3295 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 1s - loss: 1.2092 - val_loss: 24.7564 - 1s/epoch - 6ms/step
Epoch 11/20
226/226 - 2s - loss: 1.7698 - val_loss: 28.4375 - 2s/epoch - 8ms/step
Epoch 12/20
226/226 - 2s - loss: 4.3547 - val_loss: 21.3643 - 2s/epoch - 7ms/step
Epoch 13/20
226/226 - 2s - loss: 6.2425 - val_loss: 16.0654 - 2s/epoch - 9ms/step
Epoch 14/20
226/226 - 2s - loss: 3.6671 - val_loss: 16.9212 - 2s/epoch - 8ms/step
Epoch 15/20
226/226 - 2s - loss: 3.5503 - val_loss: 12.9283 - 2s/epoch - 8ms/step
Epoch 16/20
226/226 - 2s - loss: 3.5513 - val_loss: 12.0035 - 2s/epoch - 8ms/step
Epoch 17/20
226/226 - 2s - loss: 3.1212 - val_loss: 13.2702 - 2s/epoch - 7ms/step
Epoch 18/20
226/226 - 2s - loss: 2.6468 - val_loss: 13.0565 - 2s/epoch - 7ms/step
Epoch 19/20
226/226 - 2s - loss: 2.4637 - val_loss: 12.0501 - 2s/epoch - 7ms/step
Epoch 20/20
226/226 - 2s - loss: 2.2244 - val_loss: 12.2772 - 2s/epoch - 7ms/step
Run completed: runs/2024-05-06T15-02-47Z
Training run 17/20 (flags = list(0.001, 128, 32, 0.5, "relu"))
Using run directory runs/2024-05-06T15-03-24Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:03:26.549386: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0054s vs `on_train_batch_end` time: 0.0097s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0054s vs `on_train_batch_end` time: 0.0097s). Check your callbacks.
2024-05-06 10:03:28.721349: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.226/226 - 4s - loss: 7.5747 - val_loss: 3.4870 - 4s/epoch - 17ms/step
Epoch 2/20
226/226 - 2s - loss: 0.8784 - val_loss: 3.1515 - 2s/epoch - 10ms/step
Epoch 3/20
226/226 - 2s - loss: 0.4849 - val_loss: 2.5994 - 2s/epoch - 9ms/step
Epoch 4/20
226/226 - 2s - loss: 0.3130 - val_loss: 2.5727 - 2s/epoch - 8ms/step
Epoch 5/20
226/226 - 2s - loss: 0.2393 - val_loss: 2.9651 - 2s/epoch - 8ms/step
Epoch 6/20
226/226 - 1s - loss: 0.1816 - val_loss: 2.4316 - 1s/epoch - 7ms/step
Epoch 7/20
226/226 - 2s - loss: 0.1522 - val_loss: 2.4291 - 2s/epoch - 9ms/step
Epoch 8/20
226/226 - 2s - loss: 0.1375 - val_loss: 2.3906 - 2s/epoch - 7ms/step
Epoch 9/20
226/226 - 2s - loss: 0.1426 - val_loss: 2.6801 - 2s/epoch - 7ms/step
Epoch 10/20
226/226 - 2s - loss: 0.1421 - val_loss: 2.3931 - 2s/epoch - 8ms/step
Epoch 11/20
226/226 - 2s - loss: 0.1620 - val_loss: 1.9415 - 2s/epoch - 8ms/step
Epoch 12/20
226/226 - 2s - loss: 0.1833 - val_loss: 1.9059 - 2s/epoch - 8ms/step
Epoch 13/20
226/226 - 2s - loss: 0.5284 - val_loss: 4.3917 - 2s/epoch - 8ms/step
Epoch 14/20
226/226 - 2s - loss: 1.4037 - val_loss: 2.5431 - 2s/epoch - 7ms/step
Epoch 15/20
226/226 - 1s - loss: 2.9297 - val_loss: 5.9466 - 1s/epoch - 7ms/step
Epoch 16/20
226/226 - 2s - loss: 1.7710 - val_loss: 26.1538 - 2s/epoch - 8ms/step
Epoch 17/20
226/226 - 1s - loss: 2.7779 - val_loss: 19.3276 - 1s/epoch - 7ms/step
Epoch 18/20
226/226 - 2s - loss: 1.8544 - val_loss: 4.2559 - 2s/epoch - 8ms/step
Epoch 19/20
226/226 - 2s - loss: 9.7696 - val_loss: 23.6507 - 2s/epoch - 8ms/step
Epoch 20/20
226/226 - 1s - loss: 0.5702 - val_loss: 36.0097 - 1s/epoch - 7ms/step
Run completed: runs/2024-05-06T15-03-24Z
Training run 18/20 (flags = list(0.5, 32, 64, 0.3, "relu"))
Using run directory runs/2024-05-06T15-04-02Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:04:03.283512: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0045s vs `on_train_batch_end` time: 0.0089s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0045s vs `on_train_batch_end` time: 0.0089s). Check your callbacks.
2024-05-06 10:04:06.284323: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.113/113 - 4s - loss: 17.6082 - val_loss: 2.3672 - 4s/epoch - 34ms/step
Epoch 2/20
113/113 - 1s - loss: 6.9668 - val_loss: 1.7528 - 882ms/epoch - 8ms/step
Epoch 3/20
113/113 - 1s - loss: 4.7502 - val_loss: 1.8603 - 711ms/epoch - 6ms/step
Epoch 4/20
113/113 - 1s - loss: 3.1467 - val_loss: 1.7429 - 961ms/epoch - 9ms/step
Epoch 5/20
113/113 - 1s - loss: 2.2308 - val_loss: 1.9816 - 943ms/epoch - 8ms/step
Epoch 6/20
113/113 - 1s - loss: 1.6439 - val_loss: 2.2695 - 936ms/epoch - 8ms/step
Epoch 7/20
113/113 - 1s - loss: 1.1902 - val_loss: 1.7954 - 1s/epoch - 10ms/step
Epoch 8/20
113/113 - 1s - loss: 0.8874 - val_loss: 2.1697 - 707ms/epoch - 6ms/step
Epoch 9/20
113/113 - 1s - loss: 0.7588 - val_loss: 2.3111 - 934ms/epoch - 8ms/step
Epoch 10/20
113/113 - 1s - loss: 0.6286 - val_loss: 2.2440 - 1s/epoch - 12ms/step
Epoch 11/20
113/113 - 1s - loss: 0.5317 - val_loss: 2.3643 - 934ms/epoch - 8ms/step
Epoch 12/20
113/113 - 1s - loss: 0.4632 - val_loss: 2.2526 - 930ms/epoch - 8ms/step
Epoch 13/20
113/113 - 1s - loss: 0.4096 - val_loss: 2.5801 - 954ms/epoch - 8ms/step
Epoch 14/20
113/113 - 1s - loss: 0.3803 - val_loss: 2.1516 - 931ms/epoch - 8ms/step
Epoch 15/20
113/113 - 1s - loss: 0.3494 - val_loss: 2.3517 - 926ms/epoch - 8ms/step
Epoch 16/20
113/113 - 1s - loss: 0.3167 - val_loss: 2.5672 - 933ms/epoch - 8ms/step
Epoch 17/20
113/113 - 1s - loss: 0.3074 - val_loss: 2.8332 - 839ms/epoch - 7ms/step
Epoch 18/20
113/113 - 1s - loss: 0.2822 - val_loss: 2.8748 - 703ms/epoch - 6ms/step
Epoch 19/20
113/113 - 1s - loss: 0.2659 - val_loss: 2.7788 - 947ms/epoch - 8ms/step
Epoch 20/20
113/113 - 1s - loss: 0.2508 - val_loss: 2.8518 - 939ms/epoch - 8ms/step
Run completed: runs/2024-05-06T15-04-02Z
Training run 19/20 (flags = list(0.5, 32, 128, 0.3, "relu"))
Using run directory runs/2024-05-06T15-04-24Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:04:25.384704: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0049s vs `on_train_batch_end` time: 0.0095s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0049s vs `on_train_batch_end` time: 0.0095s). Check your callbacks.
2024-05-06 10:04:27.757719: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.57/57 - 3s - loss: 22.1191 - val_loss: 4.0862 - 3s/epoch - 55ms/step
Epoch 2/20
57/57 - 1s - loss: 9.2245 - val_loss: 2.9485 - 1s/epoch - 19ms/step
Epoch 3/20
57/57 - 1s - loss: 7.3146 - val_loss: 2.3437 - 861ms/epoch - 15ms/step
Epoch 4/20
57/57 - 0s - loss: 6.3196 - val_loss: 2.3771 - 424ms/epoch - 7ms/step
Epoch 5/20
57/57 - 0s - loss: 5.4482 - val_loss: 2.4674 - 431ms/epoch - 8ms/step
Epoch 6/20
57/57 - 0s - loss: 4.7752 - val_loss: 2.8429 - 449ms/epoch - 8ms/step
Epoch 7/20
57/57 - 0s - loss: 4.0931 - val_loss: 2.9437 - 432ms/epoch - 8ms/step
Epoch 8/20
57/57 - 1s - loss: 3.6478 - val_loss: 3.2997 - 1s/epoch - 19ms/step
Epoch 9/20
57/57 - 1s - loss: 3.0678 - val_loss: 3.6161 - 676ms/epoch - 12ms/step
Epoch 10/20
57/57 - 0s - loss: 2.6597 - val_loss: 3.8673 - 426ms/epoch - 7ms/step
Epoch 11/20
57/57 - 1s - loss: 2.2568 - val_loss: 4.8779 - 613ms/epoch - 11ms/step
Epoch 12/20
57/57 - 0s - loss: 1.9041 - val_loss: 4.4440 - 468ms/epoch - 8ms/step
Epoch 13/20
57/57 - 1s - loss: 1.6503 - val_loss: 5.6607 - 634ms/epoch - 11ms/step
Epoch 14/20
57/57 - 0s - loss: 1.5280 - val_loss: 6.0256 - 428ms/epoch - 8ms/step
Epoch 15/20
57/57 - 1s - loss: 1.3766 - val_loss: 5.7559 - 625ms/epoch - 11ms/step
Epoch 16/20
57/57 - 1s - loss: 1.2884 - val_loss: 6.3102 - 578ms/epoch - 10ms/step
Epoch 17/20
57/57 - 0s - loss: 1.2674 - val_loss: 5.9422 - 445ms/epoch - 8ms/step
Epoch 18/20
57/57 - 1s - loss: 1.1961 - val_loss: 5.8227 - 851ms/epoch - 15ms/step
Epoch 19/20
57/57 - 0s - loss: 1.1268 - val_loss: 6.9040 - 429ms/epoch - 8ms/step
Epoch 20/20
57/57 - 1s - loss: 1.0341 - val_loss: 6.4695 - 682ms/epoch - 12ms/step
Run completed: runs/2024-05-06T15-04-24Z
Training run 20/20 (flags = list(0.01, 64, 16, 0.2, "relu"))
Using run directory runs/2024-05-06T15-04-40Z
> FLAGS<- flags(
+ flag_numeric("nodes", 32),
+ flag_numeric("batch_size",32),
+ flag_string("activation","relu"),
+ flag_numeric("learning_ra ..." ... [TRUNCATED]
> model = keras_model_sequential()
> model %>%
+ layer_dense(units = FLAGS$nodes, activation = FLAGS$activation, input_shape = dim(carbonTrainingFinal)[2]) %>%
+ layer_dropout(rate= .... [TRUNCATED]
> model %>% compile(
+ loss="mse",
+ optimizer=optimizer_adam(lr=FLAGS$learning_rate)
+ )WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.
> model %>%fit( as.matrix(carbonTrainingFinal),
+ carbonTrainingLabels,
+ batch_size=FLAGS$batch_size,
+ .... [TRUNCATED]
Epoch 1/202024-05-06 10:04:43.116126: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0055s vs `on_train_batch_end` time: 0.0096s). Check your callbacks.
WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0055s vs `on_train_batch_end` time: 0.0096s). Check your callbacks.
2024-05-06 10:04:45.725073: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.451/451 - 6s - loss: 4.0625 - val_loss: 1.2168 - 6s/epoch - 13ms/step
Epoch 2/20
451/451 - 3s - loss: 0.3220 - val_loss: 1.2219 - 3s/epoch - 7ms/step
Epoch 3/20
451/451 - 3s - loss: 0.1863 - val_loss: 0.8444 - 3s/epoch - 7ms/step
Epoch 4/20
451/451 - 3s - loss: 0.1310 - val_loss: 1.0633 - 3s/epoch - 6ms/step
Epoch 5/20
451/451 - 3s - loss: 0.0993 - val_loss: 1.0770 - 3s/epoch - 7ms/step
Epoch 6/20
451/451 - 3s - loss: 0.0809 - val_loss: 0.9941 - 3s/epoch - 7ms/step
Epoch 7/20
451/451 - 3s - loss: 0.0737 - val_loss: 1.3372 - 3s/epoch - 7ms/step
Epoch 8/20
451/451 - 3s - loss: 0.0978 - val_loss: 0.9655 - 3s/epoch - 7ms/step
Epoch 9/20
451/451 - 3s - loss: 0.2461 - val_loss: 3.0716 - 3s/epoch - 6ms/step
Epoch 10/20
451/451 - 3s - loss: 0.4367 - val_loss: 6.2774 - 3s/epoch - 6ms/step
Epoch 11/20
451/451 - 3s - loss: 0.8453 - val_loss: 21.2237 - 3s/epoch - 6ms/step
Epoch 12/20
451/451 - 3s - loss: 1.3331 - val_loss: 34.7938 - 3s/epoch - 6ms/step
Epoch 13/20
451/451 - 3s - loss: 0.7451 - val_loss: 54.1062 - 3s/epoch - 6ms/step
Epoch 14/20
451/451 - 3s - loss: 1.0797 - val_loss: 82.9386 - 3s/epoch - 6ms/step
Epoch 15/20
451/451 - 3s - loss: 1.4240 - val_loss: 101.5648 - 3s/epoch - 6ms/step
Epoch 16/20
451/451 - 3s - loss: 1.7040 - val_loss: 212.4915 - 3s/epoch - 6ms/step
Epoch 17/20
451/451 - 3s - loss: 1.5936 - val_loss: 225.9463 - 3s/epoch - 6ms/step
Epoch 18/20
451/451 - 3s - loss: 2.6989 - val_loss: 299.2869 - 3s/epoch - 6ms/step
Epoch 19/20
451/451 - 3s - loss: 2.1928 - val_loss: 403.6202 - 3s/epoch - 6ms/step
Epoch 20/20
451/451 - 3s - loss: 4.2743 - val_loss: 442.2154 - 3s/epoch - 6ms/step
Run completed: runs/2024-05-06T15-04-40ZRuns
runs=runs[order(runs$metric_val_loss),]
runsData frame: 20 x 23 # ... with 10 more rows
# ... with 20 more columns:
# flag_nodes, flag_batch_size, flag_activation, flag_learning_rate, flag_dropout, epochs, epochs_completed, metrics, model, loss_function,
# optimizer, learning_rate, script, start, end, completed, output, source_code, context, typeview_run(runs$run_dir[1])Warning: incomplete final line found on '/var/folders/lw/zymjkl5d1g34b21y_8l475p80000gn/T//Rtmps93sC6/file3d375b21f744/source/carbonEmission.R'Warning: incomplete final line found on '/var/folders/lw/zymjkl5d1g34b21y_8l475p80000gn/T//Rtmps93sC6/file3d375b21f744/source/CarbonEmission.R'dim(carbonTrainingFinal)[1] 8001 71dim(carbonValidationFinal)[1] 799 71carbonTrainingFinal<-rbind(carbonTrainingFinal,carbonValidationFinal)
carbonTrainingLabels<-c(carbonTrainingLabels,carbonValidationLabels)
dim(carbonTrainingFinal)[1] 8800 71BestModel<-keras_model_sequential()%>%
layer_dense(units = 64,activation = "relu",input_shape = dim(carbonTrainingFinal)[2])%>%
layer_dropout(rate=0.1)%>%
layer_dense(units = 64,activation = "relu")%>%
layer_dropout(rate=0.1)%>%
layer_dense(units = 64,activation = "relu")%>%
layer_dropout(rate=0.1)%>%
layer_dense(units = 1)
BestModel %>% compile(
loss="mse",
optimizer=optimizer_adam(lr=0.001)
)WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.
WARNING:absl:There is a known slowdown when using v2.11+ Keras optimizers on M1/M2 Macs. Falling back to the legacy Keras optimizer, i.e., `tf.keras.optimizers.legacy.Adam`.history<-BestModel %>% fit(as.matrix(carbonTrainingFinal),
carbonTrainingLabels,
batch_size=128,
epochs=20,
validation_data=list(as.matrix(carbonTestingFinal),carbonTestingLabels)
)Epoch 1/202024-05-06 10:06:49.315203: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
1/69 [..............................] - ETA: 27s - loss: 67.1868
10/69 [===>..........................] - ETA: 0s - loss: 50.4680
21/69 [========>.....................] - ETA: 0s - loss: 34.5585
32/69 [============>.................] - ETA: 0s - loss: 24.4306
43/69 [=================>............] - ETA: 0s - loss: 19.2264
54/69 [======================>.......] - ETA: 0s - loss: 15.8056
64/69 [==========================>...] - ETA: 0s - loss: 13.6703
69/69 [==============================] - 1s 6ms/step - loss: 12.87612024-05-06 10:06:50.013372: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.
69/69 [==============================] - 2s 16ms/step - loss: 12.8761 - val_loss: 0.4872
Epoch 2/20
1/69 [..............................] - ETA: 0s - loss: 1.9323
11/69 [===>..........................] - ETA: 0s - loss: 1.9930
21/69 [========>.....................] - ETA: 0s - loss: 1.8814
31/69 [============>.................] - ETA: 0s - loss: 1.8593
41/69 [================>.............] - ETA: 0s - loss: 1.8222
52/69 [=====================>........] - ETA: 0s - loss: 1.7954
63/69 [==========================>...] - ETA: 0s - loss: 1.7758
69/69 [==============================] - 0s 5ms/step - loss: 1.7630
69/69 [==============================] - 1s 7ms/step - loss: 1.7630 - val_loss: 0.3016
Epoch 3/20
1/69 [..............................] - ETA: 0s - loss: 1.5298
11/69 [===>..........................] - ETA: 0s - loss: 1.5985
22/69 [========>.....................] - ETA: 0s - loss: 1.6022
32/69 [============>.................] - ETA: 0s - loss: 1.5711
43/69 [=================>............] - ETA: 0s - loss: 1.5530
55/69 [======================>.......] - ETA: 0s - loss: 1.5071
66/69 [===========================>..] - ETA: 0s - loss: 1.4773
69/69 [==============================] - 0s 5ms/step - loss: 1.4701
69/69 [==============================] - 0s 7ms/step - loss: 1.4701 - val_loss: 0.1979
Epoch 4/20
1/69 [..............................] - ETA: 0s - loss: 1.3780
12/69 [====>.........................] - ETA: 0s - loss: 1.4164
23/69 [=========>....................] - ETA: 0s - loss: 1.3816
34/69 [=============>................] - ETA: 0s - loss: 1.3575
45/69 [==================>...........] - ETA: 0s - loss: 1.3516
56/69 [=======================>......] - ETA: 0s - loss: 1.3298
67/69 [============================>.] - ETA: 0s - loss: 1.3234
69/69 [==============================] - 0s 5ms/step - loss: 1.3209
69/69 [==============================] - 0s 7ms/step - loss: 1.3209 - val_loss: 0.1723
Epoch 5/20
1/69 [..............................] - ETA: 0s - loss: 1.5950
12/69 [====>.........................] - ETA: 0s - loss: 1.2468
24/69 [=========>....................] - ETA: 0s - loss: 1.1921
36/69 [==============>...............] - ETA: 0s - loss: 1.1731
48/69 [===================>..........] - ETA: 0s - loss: 1.1598
60/69 [=========================>....] - ETA: 0s - loss: 1.1512
69/69 [==============================] - 0s 5ms/step - loss: 1.1474
69/69 [==============================] - 0s 7ms/step - loss: 1.1474 - val_loss: 0.1564
Epoch 6/20
1/69 [..............................] - ETA: 0s - loss: 1.0523
11/69 [===>..........................] - ETA: 0s - loss: 1.1100
22/69 [========>.....................] - ETA: 0s - loss: 1.0938
33/69 [=============>................] - ETA: 0s - loss: 1.0778
45/69 [==================>...........] - ETA: 0s - loss: 1.0633
56/69 [=======================>......] - ETA: 0s - loss: 1.0653
67/69 [============================>.] - ETA: 0s - loss: 1.0524
69/69 [==============================] - 0s 5ms/step - loss: 1.0502
69/69 [==============================] - 0s 7ms/step - loss: 1.0502 - val_loss: 0.1713
Epoch 7/20
1/69 [..............................] - ETA: 0s - loss: 0.8185
11/69 [===>..........................] - ETA: 0s - loss: 0.9630
22/69 [========>.....................] - ETA: 0s - loss: 0.9563
33/69 [=============>................] - ETA: 0s - loss: 0.9584
45/69 [==================>...........] - ETA: 0s - loss: 0.9490
55/69 [======================>.......] - ETA: 0s - loss: 0.9391
66/69 [===========================>..] - ETA: 0s - loss: 0.9392
69/69 [==============================] - 0s 5ms/step - loss: 0.9343
69/69 [==============================] - 1s 8ms/step - loss: 0.9343 - val_loss: 0.1433
Epoch 8/20
1/69 [..............................] - ETA: 0s - loss: 1.0760
11/69 [===>..........................] - ETA: 0s - loss: 0.8604
23/69 [=========>....................] - ETA: 0s - loss: 0.8879
35/69 [==============>...............] - ETA: 0s - loss: 0.8838
46/69 [===================>..........] - ETA: 0s - loss: 0.8758
58/69 [========================>.....] - ETA: 0s - loss: 0.8631
69/69 [==============================] - 0s 5ms/step - loss: 0.8582
69/69 [==============================] - 0s 7ms/step - loss: 0.8582 - val_loss: 0.1558
Epoch 9/20
1/69 [..............................] - ETA: 0s - loss: 0.7988
11/69 [===>..........................] - ETA: 0s - loss: 0.8231
22/69 [========>.....................] - ETA: 0s - loss: 0.8113
33/69 [=============>................] - ETA: 0s - loss: 0.8156
44/69 [==================>...........] - ETA: 0s - loss: 0.8021
55/69 [======================>.......] - ETA: 0s - loss: 0.7914
66/69 [===========================>..] - ETA: 0s - loss: 0.7827
69/69 [==============================] - 0s 5ms/step - loss: 0.7804
69/69 [==============================] - 0s 7ms/step - loss: 0.7804 - val_loss: 0.1368
Epoch 10/20
1/69 [..............................] - ETA: 0s - loss: 0.8683
10/69 [===>..........................] - ETA: 0s - loss: 0.7987
20/69 [=======>......................] - ETA: 0s - loss: 0.7896
30/69 [============>.................] - ETA: 0s - loss: 0.7798
39/69 [===============>..............] - ETA: 0s - loss: 0.7703
49/69 [====================>.........] - ETA: 0s - loss: 0.7595
59/69 [========================>.....] - ETA: 0s - loss: 0.7550
69/69 [==============================] - 0s 5ms/step - loss: 0.7436
69/69 [==============================] - 1s 8ms/step - loss: 0.7436 - val_loss: 0.1731
Epoch 11/20
1/69 [..............................] - ETA: 0s - loss: 0.6209
10/69 [===>..........................] - ETA: 0s - loss: 0.7128
20/69 [=======>......................] - ETA: 0s - loss: 0.6910
30/69 [============>.................] - ETA: 0s - loss: 0.6880
40/69 [================>.............] - ETA: 0s - loss: 0.6841
51/69 [=====================>........] - ETA: 0s - loss: 0.6828
62/69 [=========================>....] - ETA: 0s - loss: 0.6806
69/69 [==============================] - 0s 5ms/step - loss: 0.6818
69/69 [==============================] - 1s 7ms/step - loss: 0.6818 - val_loss: 0.1619
Epoch 12/20
1/69 [..............................] - ETA: 0s - loss: 0.6595
11/69 [===>..........................] - ETA: 0s - loss: 0.6448
21/69 [========>.....................] - ETA: 0s - loss: 0.6494
32/69 [============>.................] - ETA: 0s - loss: 0.6393
43/69 [=================>............] - ETA: 0s - loss: 0.6370
54/69 [======================>.......] - ETA: 0s - loss: 0.6358
65/69 [===========================>..] - ETA: 0s - loss: 0.6356
69/69 [==============================] - 0s 5ms/step - loss: 0.6363
69/69 [==============================] - 0s 7ms/step - loss: 0.6363 - val_loss: 0.1585
Epoch 13/20
1/69 [..............................] - ETA: 0s - loss: 0.5724
11/69 [===>..........................] - ETA: 0s - loss: 0.6183
22/69 [========>.....................] - ETA: 0s - loss: 0.6164
33/69 [=============>................] - ETA: 0s - loss: 0.6284
44/69 [==================>...........] - ETA: 0s - loss: 0.6211
55/69 [======================>.......] - ETA: 0s - loss: 0.6146
66/69 [===========================>..] - ETA: 0s - loss: 0.6131
69/69 [==============================] - 0s 5ms/step - loss: 0.6141
69/69 [==============================] - 0s 7ms/step - loss: 0.6141 - val_loss: 0.1638
Epoch 14/20
1/69 [..............................] - ETA: 0s - loss: 0.4695
11/69 [===>..........................] - ETA: 0s - loss: 0.5839
22/69 [========>.....................] - ETA: 0s - loss: 0.5765
33/69 [=============>................] - ETA: 0s - loss: 0.5786
44/69 [==================>...........] - ETA: 0s - loss: 0.5766
55/69 [======================>.......] - ETA: 0s - loss: 0.5688
66/69 [===========================>..] - ETA: 0s - loss: 0.5602
69/69 [==============================] - 0s 5ms/step - loss: 0.5608
69/69 [==============================] - 0s 7ms/step - loss: 0.5608 - val_loss: 0.1593
Epoch 15/20
1/69 [..............................] - ETA: 0s - loss: 0.4251
11/69 [===>..........................] - ETA: 0s - loss: 0.5252
22/69 [========>.....................] - ETA: 0s - loss: 0.5172
33/69 [=============>................] - ETA: 0s - loss: 0.5140
44/69 [==================>...........] - ETA: 0s - loss: 0.5078
55/69 [======================>.......] - ETA: 0s - loss: 0.5041
66/69 [===========================>..] - ETA: 0s - loss: 0.4987
69/69 [==============================] - 0s 5ms/step - loss: 0.4981
69/69 [==============================] - 0s 7ms/step - loss: 0.4981 - val_loss: 0.1620
Epoch 16/20
1/69 [..............................] - ETA: 0s - loss: 0.5678
11/69 [===>..........................] - ETA: 0s - loss: 0.4655
22/69 [========>.....................] - ETA: 0s - loss: 0.4634
33/69 [=============>................] - ETA: 0s - loss: 0.4639
44/69 [==================>...........] - ETA: 0s - loss: 0.4649
56/69 [=======================>......] - ETA: 0s - loss: 0.4679
68/69 [============================>.] - ETA: 0s - loss: 0.4645
69/69 [==============================] - 0s 5ms/step - loss: 0.4638
69/69 [==============================] - 0s 7ms/step - loss: 0.4638 - val_loss: 0.2482
Epoch 17/20
1/69 [..............................] - ETA: 0s - loss: 0.4742
5/69 [=>............................] - ETA: 0s - loss: 0.4301
15/69 [=====>........................] - ETA: 0s - loss: 0.4377
26/69 [==========>...................] - ETA: 0s - loss: 0.4242
37/69 [===============>..............] - ETA: 0s - loss: 0.4263
48/69 [===================>..........] - ETA: 0s - loss: 0.4307
59/69 [========================>.....] - ETA: 0s - loss: 0.4260
69/69 [==============================] - 0s 5ms/step - loss: 0.4257
69/69 [==============================] - 1s 8ms/step - loss: 0.4257 - val_loss: 0.2002
Epoch 18/20
1/69 [..............................] - ETA: 0s - loss: 0.4296
10/69 [===>..........................] - ETA: 0s - loss: 0.4093
20/69 [=======>......................] - ETA: 0s - loss: 0.3958
30/69 [============>.................] - ETA: 0s - loss: 0.3844
40/69 [================>.............] - ETA: 0s - loss: 0.3827
50/69 [====================>.........] - ETA: 0s - loss: 0.3863
60/69 [=========================>....] - ETA: 0s - loss: 0.3824
69/69 [==============================] - 0s 5ms/step - loss: 0.3816
69/69 [==============================] - 1s 8ms/step - loss: 0.3816 - val_loss: 0.2666
Epoch 19/20
1/69 [..............................] - ETA: 0s - loss: 0.3391
11/69 [===>..........................] - ETA: 0s - loss: 0.3784
22/69 [========>.....................] - ETA: 0s - loss: 0.3767
33/69 [=============>................] - ETA: 0s - loss: 0.3664
44/69 [==================>...........] - ETA: 0s - loss: 0.3631
54/69 [======================>.......] - ETA: 0s - loss: 0.3566
64/69 [==========================>...] - ETA: 0s - loss: 0.3521
69/69 [==============================] - 0s 5ms/step - loss: 0.3519
69/69 [==============================] - 1s 8ms/step - loss: 0.3519 - val_loss: 0.2001
Epoch 20/20
1/69 [..............................] - ETA: 0s - loss: 0.3680
10/69 [===>..........................] - ETA: 0s - loss: 0.3497
20/69 [=======>......................] - ETA: 0s - loss: 0.3464
30/69 [============>.................] - ETA: 0s - loss: 0.3501
40/69 [================>.............] - ETA: 0s - loss: 0.3428
50/69 [====================>.........] - ETA: 0s - loss: 0.3385
60/69 [=========================>....] - ETA: 0s - loss: 0.3350
69/69 [==============================] - 0s 5ms/step - loss: 0.3350
69/69 [==============================] - 1s 8ms/step - loss: 0.3350 - val_loss: 0.2561predictBestModel<-model %>% predict(as.matrix(carbonTestingFinal))
1/63 [..............................] - ETA: 0s
36/63 [================>.............] - ETA: 0s
63/63 [==============================] - 0s 1ms/step
63/63 [==============================] - 0s 1ms/steprmse=function(x,y){
return((mean(x-y)^2)^0.5)
}
rmse(predictBestModel,carbonTestingLabels)[1] 0.2110004MAE(predictBestModel,carbonTestingLabels)[1] 0.2486037rsquaredBest<-sum((predictBestModel-carbonTestingLabels)^2)/sum((carbonTestingLabels-mean(carbonTestingLabels))^2)
rsquaredBest[1] 0.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